Neo4j-Manual-2 0 0

The Neo4j Manual v2.0.
The Neo4j Team neo4j.org <http://neo4j.org/>

www.neotechnology.com <http://www.neotechnology.com/>
The Neo4j Manual v2.0.0
by The Neo4j Team neo4j.org <http://neo4j.org/> www.neotechnology.com <http://www.neotechnology.com/>

Publication date 2013-12-0815:48:51
Copyright 2013 Neo Technology
Starting points
What is a graph database?

Cypher Query Language
Languages / Remote Client Libraries
REST API
Installation
Upgrading
Security
License: Creative Commons 3.0
This book is presented in open source and licensed through Creative Commons 3.0. You are free to copy, distribute, transmit, and/or adapt the work. This
license is based upon the following conditions:
Attribution. You must attribute the work in the manner specified by the author or licensor (but not in any way that suggests that they endorse you or your
use of the work).
Share Alike. If you alter, transform, or build upon this work, you may distribute the resulting work only under the same, similar or a compatible license.
Any of the above conditions can be waived if you get permission from the copyright holder.
In no way are any of the following rights affected by the license:
Your fair dealing or fair use rights
The authors moral rights
Rights other persons may have either in the work itself or in how the work is used, such as publicity or privacy rights
Note
For any reuse or distribution, you must make clear to the others the license terms of this work. The best way to do this is with a direct link to
this page: http://creativecommons.org/licenses/by-sa/3.0/ <http://creativecommons.org/licenses/by-sa/3.0/>
Table of Contents
Preface .................................................................................................................................................... v
I. Introduction ........................................................................................................................................ 1
1. Neo4j Highlights ....................................................................................................................... 2
2. Graph Database Concepts ......................................................................................................... 3
3. The Neo4j Graph Database ..................................................................................................... 11
II. Tutorials .......................................................................................................................................... 21
4. Getting started with Cypher .................................................................................................... 22
5. Data Modeling Examples ........................................................................................................ 34
6. Languages ................................................................................................................................ 77
III. Cypher Query Language ............................................................................................................... 83
7. Introduction .............................................................................................................................. 84
8. Syntax ...................................................................................................................................... 97
9. General Clauses ..................................................................................................................... 113
10. Reading Clauses .................................................................................................................. 129
11. Writing Clauses ................................................................................................................... 157
12. Functions .............................................................................................................................. 180
13. Schema ................................................................................................................................. 207
14. From SQL to Cypher .......................................................................................................... 211
IV. Reference ..................................................................................................................................... 218
15. Capabilities .......................................................................................................................... 219
16. Transaction Management .................................................................................................... 226
17. Data Import .......................................................................................................................... 234
18. Graph Algorithms ................................................................................................................ 235
19. REST API ............................................................................................................................ 237
20. Deprecations ........................................................................................................................ 356
V. Operations ..................................................................................................................................... 357
21. Installation & Deployment .................................................................................................. 358
22. Configuration & Performance ............................................................................................. 372
23. High Availability ................................................................................................................. 407
24. Backup ................................................................................................................................. 431
25. Security ................................................................................................................................ 437
26. Monitoring ........................................................................................................................... 443
VI. Tools ............................................................................................................................................ 459
27. Web Interface ...................................................................................................................... 460
28. Neo4j Shell .......................................................................................................................... 461
VII. Community ................................................................................................................................. 478
29. Community Support ............................................................................................................ 479
30. Contributing to Neo4j .......................................................................................................... 480
VIII. Advanced Usage ....................................................................................................................... 504
31. Extending the Neo4j Server ................................................................................................ 505
32. Using Neo4j embedded in Java applications ...................................................................... 513
33. The Traversal Framework ................................................................................................... 545
34. Legacy Indexing .................................................................................................................. 554
35. Batch Insertion ..................................................................................................................... 573
A. Manpages ...................................................................................................................................... 577
neo4j ........................................................................................................................................... 578
neo4j-installer ............................................................................................................................. 580
iii
The Neo4j Manual v2.0.0

neo4j-shell .................................................................................................................................. 581
neo4j-backup .............................................................................................................................. 583
neo4j-arbiter ............................................................................................................................... 585
iv
Preface
This is the reference manual for Neo4j version 2.0.0, authored by the Neo4j Team.
The main parts of the manual are:
PartI, Introduction introducing graph database concepts and Neo4j.

PartII, Tutorials learn how to use Neo4j.
PartIII, Cypher Query Language details on the Cypher query language.
PartIV, Reference detailed information on Neo4j.
PartV, Operations how to install and maintain Neo4j.
PartVI, Tools guides on tools.
PartVII, Community getting help from, contributing to.
PartVIII, Advanced Usage using Neo4j in more advanced ways.
AppendixA, Manpages command line documentation.
The material is practical, technical, and focused on answering specific questions. It addresses how
things work, what to do and what to avoid to successfully run Neo4j in a production environment.
The goal is to be thumb-through and rule-of-thumb friendly.
Each section should stand on its own, so you can hop right to whatever interests you. When possible,
the sections distill rules of thumb which you can keep in mind whenever you wander out of the
house without this manual in your back pocket.
The included code examples are executed when Neo4j is built and tested. Also, the REST API request
and response examples are captured from real interaction with a Neo4j server. Thus, the examples are
always in sync with how Neo4j actually works.
Theres other documentation resources besides the manual as well:
Neo4j Cypher Refcard, see http://docs.neo4j.org/ for available versions.
Neo4j GraphGist, an online tool for creating interactive web pages with executable Cypher
statements: http://gist.neo4j.org/.
The main Neo4j site at http://www.neo4j.org/ is a good starting point to learn about Neo4j.
Who should read this?
The topics should be relevant to architects, administrators, developers and operations personnel.
PartI.Introduction
This part gives a birds eye view of what a graph database is, and then outlines some specifics of Neo4j.
Chapter1.Neo4j Highlights
As a robust, scalable and high-performance database, Neo4j is suitable for full enterprise deployment
or a subset of the full server can be used in lightweight projects.
It features:
true ACID transactions,

high availability,
scales to billions of nodes and relationships,
high speed querying through traversals,
declarative graph query language.
Proper ACID behavior is the foundation of data reliability. Neo4j enforces that all operations that
modify data occur within a transaction, guaranteeing consistent data. This robustness extends from
single instance embedded graphs to multi-server high availability installations. For details, see
Chapter16, Transaction Management.
Reliable graph storage can easily be added to any application. A graph can scale in size and
complexity as the application evolves, with little impact on performance. Whether starting new
development, or augmenting existing functionality, Neo4j is only limited by physical hardware.
A single server instance can handle a graph of billions of nodes and relationships. When data
throughput is insufficient, the graph database can be distributed among multiple servers in a high
availability configuration. See Chapter23, High Availability to learn more.
The graph database storage shines when storing richly-connected data. Querying is performed through
traversals, which can perform millions of traversal steps per second. A traversal step resembles a join
in a RDBMS.
Chapter2.Graph Database Concepts

This chapter contains an introduction to the graph data model and also compares it to other data
models used when persisting data.
Graph Database Concepts
2.1.What is a Graph Database?

A graph database stores data in a graph, the most generic of data structures, capable of elegantly
representing any kind of data in a highly accessible way. Lets follow along some graphs, using them
to express graph concepts. Well read a graph by following arrows around the diagram to form
sentences.
2.1.1.A Graph contains Nodes and Relationships

A Graph records data in Nodes which have Properties
The simplest possible graph is a single Node, a record that has named values referred to as Properties.
A Node could start with a single Property and grow to a few million Properties, though that can get a
little awkward. At some point it makes sense to distribute the data into multiple nodes, organized with
explicit Relationships.
Graph
records dat a in
records dat a in
Labels
Relat ionships
group
organize
Nodes
have
have
Propert ies
2.1.2.Relationships organize the Graph

Nodes are organized by Relationships which also have Properties
Relationships organize Nodes into arbitrary structures, allowing a Graph to resemble a List, a Tree,
a Map, or a compound Entity any of which can be combined into yet more complex, richly interconnected structures.
2.1.3.Labels group the Nodes

Nodes are grouped by Labels into Sets
Labels are a means of grouping the nodes in the graph. They can be used to restrict queries to subsets
of the graph, as well as enabling optional model constraints and indexing rules.
4
2.1.4.Query a Graph with a Traversal

A Traversal navigates a Graph; it identifies Paths which order Nodes
A Traversal is how you query a Graph, navigating from starting Nodes to related Nodes according to
an algorithm, finding answers to questions like what music do my friends like that I dont yet own,
or if this power supply goes down, what web services are affected?
Traversal
navigat es
ident ifies
Graph
Algorit hm
records dat a in
Relat ionships
expresses
Pat hs
records dat a in
order
organize
Nodes
2.1.5.Indexes look-up Nodes or Relationships

An Index maps from Properties to either Nodes or Relationships
Often, you want to find a specific Node or Relationship according to a Property it has. Rather than
traversing the entire graph, use an Index to perform a look-up, for questions like find the Account for
username master-of-graphs.
Indexes
m ap t o
Relat ionships
m ap t o
organize
have
m ap from
Nodes
have
Propert ies
2.1.6.Neo4j is a Graph Database

A Graph Database manages a Graph and also manages related Indexes
Neo4j is a commercially supported open-source graph database. It was designed and built from the
ground-up to be a reliable database, optimized for graph structures instead of tables. Working with
Neo4j, your application gets all the expressiveness of a graph, with all the dependability you expect
out of a database.
Graph Dat abase
Traversal
m anages
m anages
Indexes
navigat es
ident ifies
Graph
m ap t o
m ap t o
Algorit hm
records dat a in
Relat ionships
m ap from
records dat a in
organize
have
expresses
Nodes
have
Propert ies
Pat hs
order
2.2.Comparing Database Models

A Graph Database stores data structured in the Nodes and Relationships of a graph. How does this
compare to other persistence models? Because a graph is a generic structure, lets compare how a few
models would look in a graph.
2.2.1.A Graph Database transforms a RDBMS

Topple the stacks of records in a relational database while keeping all the relationships, and youll see
a graph. Where an RDBMS is optimized for aggregated data, Neo4j is optimized for highly connected
data.
Figure2.1.RDBMS
A1
B1
A2
B2
A3
B3
B4
B5
C1
B6
C2
B7
C3
Figure2.2.Graph Database as RDBMS

C3
A1
A2
C1
B2
B6
B1
B4
B5
A3
C2
B3
B7
2.2.2.A Graph Database elaborates a Key-Value Store

A Key-Value model is great for lookups of simple values or lists. When the values are themselves
interconnected, youve got a graph. Neo4j lets you elaborate the simple data structures into more
complex, interconnected data.

Figure2.3.Key-Value Store
V1
K2
K1
K2
V2
K3
K1
K3
V3
K1
K*
represents a key, V* a value. Note that some keys point to other keys as well as plain values.
Figure2.4.Graph Database as Key-Value Store
K2
K1
V2
V1
V3
K3
2.2.3.A Graph Database relates Column-Family

Column Family (BigTable-style) databases are an evolution of key-value, using "families" to allow
grouping of rows. Stored in a graph, the families could become hierarchical, and the relationships
among data becomes explicit.
2.2.4.A Graph Database navigates a Document Store

The container hierarchy of a document database accommodates nice, schema-free data that can easily
be represented as a tree. Which is of course a graph. Refer to other documents (or document elements)
within that tree and you have a more expressive representation of the same data. When in Neo4j, those
relationships are easily navigable.

Figure2.5.Document Store
D1
D2
S1
D2/S2
S3
V1
D1/S1
D=Document, S=Subdocument, V=Value, D2/S2
S2
V4
V3
V2
= reference to subdocument in (other) document.
Figure2.6.Graph Database as Document Store

V1
D1
V2
D2
S3
S1
S2
V3
V4
10
Chapter3.The Neo4j Graph Database

This chapter goes into more detail on the data model and behavior of Neo4j.
11
The Neo4j Graph Database
3.1.Nodes
The fundamental units that form a graph are nodes and relationships. In Neo4j, both nodes and
relationships can contain properties.
Nodes are often used to represent entities, but depending on the domain relationships may be used for
that purpose as well.
Apart from properties and relationships, nodes can also be labeled with zero or more labels.
A Node
can have
can have
Relat ionships
can have
Labels
can have
Propert ies
Lets start out with a really simple graph, containing only a single node with one property:
nam e: Pet er
12
3.2.Relationships
Relationships between nodes are a key part of a graph database. They allow for finding related data.
Just like nodes, relationships can have properties.
A Relat ionship
has a
St art node
has a
has a
End node
can have
Relat ionship t ype
Propert ies
uniquely ident ified by
Nam e
A relationship connects two nodes, and is guaranteed to have valid start and end nodes.
St art node
relat ionship
End node
As relationships are always directed, they can be viewed as outgoing or incoming relative to a node,
which is useful when traversing the graph:
incom ing relat ionship
Node
out going relat ionship
Relationships are equally well traversed in either direction. This means that there is no need to add
duplicate relationships in the opposite direction (with regard to traversal or performance).
While relationships always have a direction, you can ignore the direction where it is not useful in your
application.
Note that a node can have relationships to itself as well:
Node
loop
To further enhance graph traversal all relationships have a relationship type. Note that the word type
might be misleading here, you could rather think of it as a label. The following example shows a
simple social network with two relationship types.
13
Maja
follows
Alice
follows
follows
Oscar
blocks
William
Using relationship direction and type

What
How
get who a person follows
outgoing follows relationships, depth one
get the followers of a person
incoming follows relationships, depth one
get who a person blocks
outgoing blocks relationships, depth one
get who a person is blocked by
incoming blocks relationships, depth one
14
3.3.Properties
Both nodes and relationships can have properties.
Properties are key-value pairs where the key is a string. Property values can be either a primitive or an
array of one primitive type. For example String, int and int[] values are valid for properties.
Note
NULL
is not a valid property value. NULLs can instead be modeled by the absence of a key.
A Propert y
has a
Value
has a
Key
can be an array of can be a
is a
boolean
byt e
short
int
Prim it ive
long
float
double
char
St ring
Property value types

Type
Description
Value range
true/false
boolean
byte
8-bit integer
-128
short
16-bit integer
-32768
int
32-bit integer
-2147483648
long
64-bit integer
-9223372036854775808 to
9223372036854775807, inclusive
float
32-bit IEEE 754 floating-point number
double
64-bit IEEE 754 floating-point number

15
to 127, inclusive
to 32767, inclusive
to 2147483647, inclusive

Type
Description
Value range
char
16-bit unsigned integers representing

Unicode characters
u0000
String
sequence of Unicode characters
to uffff (0 to 65535)
For further details on float/double values, see Java Language Specification <http://docs.oracle.com/
javase/specs/jls/se5.0/html/typesValues.html#4.2.3>.
16
3.4.Labels
A label is a named graph construct that is used to group nodes into sets; all nodes labeled with the
same label belongs to the same set. Many database queries can work with these sets instead of the
whole graph, making queries easier to write and more efficient. A node may be labeled with any
number of labels, including none, making labels an optional addition to the graph.
A Label
has a
Nam e
groups
Node
Labels are used when defining contraints and adding indexes for properties.
An example would be a label named User that you label all your nodes representing users with. With
that in place, you can ask Neo4j to perform operations only on your user nodes, such as finding all
users with a given name.
However, you can use labels for much more. For instance, since labels can be added and removed
during runtime, they can be used to mark temporary states for your nodes. You might create an
Offline label for phones that are offline, a Happy label for happy pets, and so on.
3.4.1.Label names
Any non-empty unicode string can be used as a label name. In Cypher, you may need to use the
backtick (`) syntax to avoid clashes with Cypher identifier rules. By convention, labels are written
with CamelCase notation, with the first letter in upper case. For instance, User or CarOwner.
Labels have an id space of an int, meaning the maximum number of labels the database can contain is
roughly 2 billion.
17
3.5.Paths
A path is one or more nodes with connecting relationships, typically retrieved as a query or traversal
result.
A Pat h
has a can cont ain one or m ore
St art Node
Relat ionship
accom panied by a
Node
The shortest possible path has length zero and looks like this:
Node
A path of length one:

Node 1
Relat ionship 1
Node 2
Another path of length one:

Node 1
Relat ionship 1
18
has an
End Node
3.6.Traversal
Traversing a graph means visiting its nodes, following relationships according to some rules. In most
cases only a subgraph is visited, as you already know where in the graph the interesting nodes and
relationships are found.
Cypher provides a declarative way to query the graph powered by traversals and other techniques. See
PartIII, Cypher Query Language for more information.
Neo4j comes with a callback based traversal API which lets you specify the traversal rules. At a basic
level theres a choice between traversing breadth- or depth-first.
For an in-depth introduction to the traversal framework, see Chapter33, The Traversal Framework.
For Java code examples see Section32.7, Traversal.
19
3.7.Schema
Neo4j is a schema-optional graph database. You can use Neo4j without any schema. Optionally you
can introduce it in order to gain performance or modeling benefits. This allows a way of working
where the schema does not get in your way until you are at a stage where you want to reap the benefits
of having one.
3.7.1.Indexes
Note
This feature was introduced in Neo4j 2.0, and is not the same as the legacy indexes (see
Chapter34, Legacy Indexing).
Performance is gained by creating indexes, which improve the speed of looking up nodes in the
database. Once youve specified which properties to index, Neo4j will make sure your indexes are
kept up to date as your graph evolves. Any operation that looks up nodes by the newly indexed
properties will see a significant performance boost.
Indexes in Neo4j are eventually available. That means that when you first create an index, the
operation returns immediately. The index is populating in the background and so is not immediately
available for querying. When the index has been fully populated it will eventually come online. That
means that it is now ready to be used in queries.
If something should go wrong with the index, it can end up in a failed state. When it is failed, it will
not be used to speed up queries. To rebuild it, you can drop and recreate the index. Look at logs for
clues about the failure.
You can track the status of your index by asking for the index state through the API you are using.
Note, however, that this is not yet possible through Cypher.
How to use indexes in the different APIs:
Cypher: Section13.1, Indexes

REST API: Section19.13, Indexing
Listing Indexes via Shell: Section28.6.11, Listing Indexes and Constraints
Java Core API: Section32.4, User database with indexes
3.7.2.Constraints
Note
This feature was introduced in Neo4j 2.0.

Neo4j can help you keep your data clean. It does so using constraints, that allow you to specify the
rules for what your data should look like. Any changes that break these rules will be denied.
In this version, unique constraints is the only available constraint type.
How to use constraints in the different APIs:
Cypher: Section13.2, Constraints
REST API: Section19.14, Constraints
Listing Constraints via Shell: Section28.6.11, Listing Indexes and Constraints
20
PartII.Tutorials
The tutorial part describes how use Neo4j. It takes you from Hello World to advanced usage of graphs.
Chapter4.Getting started with Cypher

This chapter will guide you through your first steps with Cypher.
In the online edition of this manual, all queries in this section can be executed interactively without
installing Neo4j on your computer.
Otherwise, first get the Neo4j server running to try things out locally. Instructions are found in
Section21.2, Server Installation. With the server running, you can choose to issue Cypher queries
from either the web interface or the Neo4j shell. See Chapter27, Web Interface or Chapter28, Neo4j
Shell.
22
Getting started with Cypher
4.1.Create nodes and relationships

Create a node for the actor Tom Hanks:
CREATE (n:Actor { name:"Tom Hanks" });
Lets find the node we created:

MATCH (actor:Actor { name: "Tom Hanks" })
RETURN actor;
Now lets create a movie and connect it to the Tom Hanks node with an ACTED_IN relationship:
MATCH (actor:Actor)
WHERE actor.name = "Tom Hanks"
CREATE (movie:Movie { title:'Sleepless IN Seattle' })
CREATE (actor)-[:ACTED_IN]->(movie);
Using a WHERE clause in the query above to get the Tom Hanks node does the same thing as the pattern
in the MATCH clause of the previous query.
This is how our graph looks now:
Act or
nam e = 'Tom Hanks'
ACTED_IN
Movie
t it le = 'Sleepless in Seat t le'
We can do more of the work in a single clause. CREATE UNIQUE will make sure we dont create duplicate
patterns. Using this: [r:ACTED_IN] lets us return the relationship.
CREATE UNIQUE (actor)-[r:ACTED_IN]->(movie:Movie { title:"Forrest Gump" })
RETURN r;
Set a property on a node:

SET actor.DoB = 1944
RETURN actor.name, actor.DoB;
The labels Actor and Movie help us organize the graph. Lets list all Movie nodes:
MATCH (movie:Movie)
RETURN movie AS Àll Movies`;
All Movies
Node[1]{title:"Sleepless in Seattle"}
Node[2]{title:"Forrest Gump"}
2 rows
23
4.2.Movie Database
Our example graph consists of movies with title and year and actors with a name. Actors have ACTS_IN
relationships to movies, which represents the role they played. This relationship also has a role
attribute.
Well go with three movies and three actors:
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
(matrix1:Movie { title : 'The Matrix', year : '1999-03-31' })

(matrix2:Movie { title : 'The Matrix Reloaded', year : '2003-05-07' })
(matrix3:Movie { title : 'The Matrix Revolutions', year : '2003-10-27' })
(keanu:Actor { name:'Keanu Reeves' })
(laurence:Actor { name:'Laurence Fishburne' })
(carrieanne:Actor { name:'Carrie-Anne Moss' })
(keanu)-[:ACTS_IN { role : 'Neo' }]->(matrix1)
(laurence)-[:ACTS_IN { role : 'Morpheus' }]->(matrix1)
(carrieanne)-[:ACTS_IN { role : 'Trinity' }]->(matrix1)
This gives us the following graph to play with:

Act or
nam e = 'Keanu Reeves'
ACTS_IN
role = 'Neo'
ACTS_IN
role = 'Trinit y'
Act or
ACTS_IN
role = 'Neo'
Movie
t it le = 'The Mat rix Reloaded'
year = '2003-05-07'
Act or
nam e = 'Carrie-Anne Moss'
ACTS_IN
role = 'Trinit y'
nam e = 'Laurence Fishburne'
ACTS_IN
role = 'Morpheus'
ACTS_IN
role = 'Neo'
Movie
ACTS_IN
role = 'Trinit y'
Movie
t it le = 'The Mat rix'

year = '1999-03-31'
t it le = 'The Mat rix Revolut ions'

year = '2003-10-27'
Lets check how many nodes we have now:

MATCH (n)
RETURN "Hello Graph with " + count(*)+ " Nodes!" AS welcome;
Return a single node, by name:

MATCH (movie:Movie { title: 'The Matrix' })
RETURN movie;
Return the title and date of the matrix node:

MATCH (movie:Movie { title: 'The Matrix' })
RETURN movie.title, movie.year;
Which results in:

movie.title
movie.year
"The Matrix"
"1999-03-31"
1 row
Show all actors:
MATCH (actor:Actor)
RETURN actor;
24
ACTS_IN
role = 'Morpheus'
ACTS_IN
role = 'Morpheus'

Return just the name, and order them by name:
MATCH (actor:Actor)
RETURN actor.name
ORDER BY actor.name;
Count the actors:

MATCH (actor:Actor)
RETURN count(*);
Get only the actors whose names end with s:

MATCH (actor:Actor)
WHERE actor.name =~ ".*s$"
RETURN actor.name;
Heres some exploratory queries for unknown datasets. Dont do this on live production databases!
Count nodes:
MATCH (n)
RETURN count(*);
Count relationship types:

MATCH (n)-[r]->()
RETURN type(r), count(*);
type(r)
count(*)
"ACTS_IN"
1 row
List all nodes and their relationships:
MATCH (n)-[r]->(m)
RETURN n AS from, r AS `->`, m AS to;
from
->
to
Node[3]{name:"Keanu Reeves"}
:ACTS_IN[0]{role:"Neo"}
Node[0]{title:"The Matrix",
year:"1999-03-31"}
Node[1]{title:"The Matrix
Reloaded", year:"2003-05-07"}
Revolutions", year:"2003-10-27"}
Node[4]{name:"Laurence
Fishburne"}
:ACTS_IN[3]{role:"Morpheus"}
year:"1999-03-31"}
Fishburne"}
Fishburne"}
Node[5]{name:"Carrie-Anne
Moss"}
:ACTS_IN[6]{role:"Trinity"}
year:"1999-03-31"}
9 rows
25

from
->
to
Moss"}
Moss"}
9 rows
26
4.3.Social Movie Database

attribute.
So far, we queried the movie data; now lets update the graph too.
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE

We will add ourselves, friends and movie ratings.

Heres how to add a node for yourself and return it, lets say your name is Me:
CREATE (me:User { name: "Me" })
RETURN me;
me
Node[6]{name:"Me"}
1 row
Nodes created: 1
Properties set: 1
Labels added: 1
Lets check if the node is there:
MATCH (me:User { name: "Me" })
RETURN me.name;
Add a movie rating:

MATCH (me:User { name: "Me" }),(movie:Movie { title: "The Matrix" })
CREATE (me)-[:RATED { stars : 5, comment : "I love that movie!" }]->(movie);
Which movies did I rate?

MATCH (me:User { name: "Me" }),(me)-[rating:RATED]->(movie)
RETURN movie.title, rating.stars, rating.comment;
movie.title
rating.stars
rating.comment
"The Matrix"
"I love that movie!"
1 row
27

We need a friend!
CREATE (friend:User { name: "A Friend" })
RETURN friend;
Add our friendship idempotently, so we can re-run the query without adding it several times. We
return the relationship to check that it has not been created several times.
MATCH (me:User { name: "Me" }),(friend:User { name: "A Friend" })
CREATE UNIQUE (me)-[friendship:FRIEND]->(friend)
RETURN friendship;
You can rerun the query, see that it doesnt change anything the second time!
Lets update our friendship with a since property:
MATCH (me:User { name: "Me" })-[friendship:FRIEND]->(friend:User { name: "A Friend" })
SET friendship.since='forever'
RETURN friendship;
Lets pretend us being our friend and wanting to see which movies our friends have rated.
MATCH (me:User { name: "A Friend" })-[:FRIEND]-(friend)-[rating:RATED]->(movie)
RETURN movie.title, avg(rating.stars) AS stars, collect(rating.comment) AS comments, count(*);
movie.title
stars
comments
count(*)
"The Matrix"
5. 0
["I love that movie!"]
1 row
Thats too little data, lets add some more friends and friendships.
MATCH (me:User { name: "Me" })
FOREACH (i IN range(1,10)| CREATE (friend:User { name: "Friend " + i }),(me)-[:FRIEND]->(friend));
Show all our friends:

MATCH (me:User { name: "Me" })-[r:FRIEND]->(friend)
RETURN type(r) AS friendship, friend.name;
friendship
friend.name
"FRIEND"
"A Friend"
"FRIEND"
"Friend 1"
"FRIEND"
"Friend 2"
"FRIEND"
"Friend 3"
"FRIEND"
"Friend 4"
"FRIEND"
"Friend 5"
"FRIEND"
"Friend 6"
"FRIEND"
"Friend 7"
"FRIEND"
"Friend 8"
"FRIEND"
"Friend 9"
"FRIEND"
"Friend 10"
11 rows
28
4.4.Finding Paths
attribute.
We queried and updated the data so far, now lets find interesting constellations, a.k.a. paths.
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE
CREATE

All other movies that actors in The Matrix acted in ordered by occurrence:
MATCH (:Movie { title: "The Matrix" })<-[:ACTS_IN]-(actor)-[:ACTS_IN]->(movie)
RETURN movie.title, count(*)
ORDER BY count(*) DESC ;
movie.title
count(*)
"The Matrix Revolutions"
"The Matrix Reloaded"
2 rows
Lets see who acted in each of these movies:
MATCH (:Movie { title: "The Matrix" })<-[:ACTS_IN]-(actor)-[:ACTS_IN]->(movie)
RETURN movie.title, collect(actor.name), count(*) AS count
ORDER BY count DESC ;
movie.title
collect(actor.name)
count
"The Matrix Revolutions"
["Keanu Reeves", "Laurence

Fishburne", "Carrie-Anne Moss"]
"The Matrix Reloaded"
["Keanu Reeves", "Laurence

Fishburne", "Carrie-Anne Moss"]
2 rows
What about co-acting, that is actors that acted together:
MATCH (:Movie { title: "The Matrix"
})<-[:ACTS_IN]-(actor)-[:ACTS_IN]->(movie)<-[:ACTS_IN]-(colleague)
RETURN actor.name, collect(DISTINCT colleague.name);
actor.name
collect(distinct colleague.name)
"Carrie-Anne Moss"
["Keanu Reeves", "Laurence Fishburne"]
3 rows
29

actor.name
collect(distinct colleague.name)
"Laurence Fishburne"
["Keanu Reeves", "Carrie-Anne Moss"]
"Keanu Reeves"
["Laurence Fishburne", "Carrie-Anne Moss"]
3 rows
Who of those other actors acted most often with anyone from the matrix cast?
MATCH (:Movie { title: "The Matrix"
})<-[:ACTS_IN]-(actor)-[:ACTS_IN]->(movie)<-[:ACTS_IN]-(colleague)
RETURN colleague.name, count(*)
ORDER BY count(*) DESC LIMIT 10;
colleague.name
count(*)
"Carrie-Anne Moss"
"Keanu Reeves"
"Laurence Fishburne"
3 rows
Starting with paths, a path is a sequence of nodes and relationships from a start node to an end node.
We know that Trinity loves Neo, but how many paths exist between their actors? Well limit the path
length and the query as it exhaustively searches the graph otherwise
MATCH p =(:Actor { name: "Keanu Reeves" })-[:ACTS_IN*0..5]-(:Actor { name: "Carrie-Anne Moss" })
RETURN p, length(p)
LIMIT 10;
length(p)
[Node[3]{name:"Keanu Reeves"}, :ACTS_IN[0]

4
{role:"Neo"}, Node[0]{title:"The
Matrix", year:"1999-03-31"}, :ACTS_IN[3]
{role:"Morpheus"}, Node[4]{name:"Laurence
Fishburne"}, :ACTS_IN[4]{role:"Morpheus"},
Node[1]{title:"The Matrix Reloaded",
year:"2003-05-07"}, :ACTS_IN[7]{role:"Trinity"},
Node[5]{name:"Carrie-Anne Moss"}]
4
{role:"Neo"}, Node[0]{title:"The
Matrix", year:"1999-03-31"}, :ACTS_IN[3]
Node[2]{title:"The Matrix Revolutions",
2
{role:"Neo"}, Node[0]{title:"The Matrix",
9 rows
30

p
length(p)

4
{role:"Neo"}, Node[1]{title:"The Matrix
Reloaded", year:"2003-05-07"}, :ACTS_IN[4]
4
Node[2]{title:"The Matrix Revolutions",
{role:"Trinity"}, Node[5]{name:"Carrie-Anne
Moss"}]

4
Revolutions", year:"2003-10-27"}, :ACTS_IN[5]
4
Node[1]{title:"The Matrix Reloaded",
{role:"Trinity"}, Node[5]{name:"Carrie-Anne
Moss"}]
9 rows
Bur thats a lot of data, we just want to look at the names and titles of the nodes of the path.
MATCH p =(:Actor { name: "Keanu Reeves" })-[:ACTS_IN*0..5]-(:Actor { name: "Carrie-Anne Moss" })
31

RETURN extract(n IN nodes(p)| coalesce(n.title,n.name)) AS `names AND titles`, length(p)
ORDER BY length(p)
LIMIT 10;
names and titles
length(p)
["Keanu Reeves", "The Matrix", "Carrie-Anne

Moss"]
["Keanu Reeves", "The Matrix Reloaded", "CarrieAnne Moss"]
["Keanu Reeves", "The Matrix

Revolutions", "Carrie-Anne Moss"]
["Keanu Reeves", "The Matrix", "Laurence

Fishburne", "The Matrix Reloaded", "Carrie-Anne
Moss"]
["Keanu Reeves", "The Matrix", "Laurence

Fishburne", "The Matrix Revolutions", "CarrieAnne Moss"]
["Keanu Reeves", "The Matrix Reloaded", "Laurence 4

Fishburne", "The Matrix", "Carrie-Anne Moss"]
["Keanu Reeves", "The Matrix Reloaded", "Laurence 4
Fishburne", "The Matrix Revolutions", "CarrieAnne Moss"]
Revolutions", "Laurence Fishburne", "The
Matrix", "Carrie-Anne Moss"]

Revolutions", "Laurence Fishburne", "The Matrix
Reloaded", "Carrie-Anne Moss"]
9 rows
32
4.5.Labels, Constraints and Indexes

Labels are a convenient way to group nodes together. They are used to restrict queries, define
constraints and create indexes.
The following will give an example of how to use labels. Lets start out adding a constraint in this
case we decided that all Movie node titles should be unique.
CREATE CONSTRAINT ON (movie:Movie) ASSERT movie.title IS UNIQUE
Note that adding the unique constraint will add an index on that property, so we wont do that
separately. If we drop the constraint, we will have to add an index instead, as needed.
In this case we want an index to speed up finding actors by name in the database:
CREATE INDEX ON :Actor(name)
Indexes can be added at any time. Constraints can be added after a label is already in use, but that
requires that the existing data complies with the constraints. Note that it will take some time for an
index to come online when theres existing data.
Now, lets add some data.
CREATE (actor:Actor { name:"Tom Hanks" }),(movie:Movie { title:'Sleepless IN Seattle' }),
(actor)-[:ACTED_IN]->(movie);
Normally you dont specify indexes when querying for data. They will be used automatically. This
means we can simply look up the Tom Hanks node, and the index will kick in behind the scenes to
boost performance.
RETURN actor;
Now lets say we want to add another label for a node. Heres how to do that:
SET actor :American;
To remove a label from nodes, this is what to do:

REMOVE actor:American;
For more information on labels and related topics, see:
Section3.4, Labels
Chapter13, Schema
Section13.2, Constraints
Section13.1, Indexes
Section9.7, Using
Section11.4, Set
Section11.6, Remove
33
Chapter5.Data Modeling Examples

The following chapters contain simplified examples of how different domains can be modeled using
Neo4j. The aim is not to give full examples, but to suggest possible ways to think using nodes,
relationships, graph patterns and data locality in traversals.
The examples use Cypher queries a lot, read PartIII, Cypher Query Language for more information.
34
Data Modeling Examples
5.1.Linked Lists
A powerful feature of using a graph database, is that you can create your own in-graph data
structures for example a linked list.
This data structure uses a single node as the list reference. The reference has an outgoing relationship
to the head of the list, and an incoming relationship from the last element of the list. If the list is
empty, the reference will point to itself.
To make it clear what happens, we will show how the graph looks after each query.
To initialize an empty linked list, we simply create a node, and make it link to itself. Unlike the actual
list elements, it doesnt have a value property.
CREATE (root { name: 'ROOT' })-[:LINK]->(root)
RETURN root
nam e = 'ROOT'
LINK
Adding values is done by finding the relationship where the new value should be placed in, and
replacing it with a new node, and two relationships to it. We also have to handle the fact that the
before and after nodes could be the same as the root node. The case where before, after and the root
node are all the same, makes it necessary to use CREATE UNIQUE to not create two new value nodes by
mistake.
MATCH (root)-[:LINK*0..]->(before),(after)-[:LINK*0..]->(root),(before)-[old:LINK]->(after)
WHERE root.name = 'ROOT' AND (before.value < 25 OR before = root) AND (25 < after.value OR after =
root)
CREATE UNIQUE (before)-[:LINK]->({ value:25 })-[:LINK]->(after)
DELETE old
nam e = 'ROOT'
LINK
LINK
value = 25
Lets add one more value:

MATCH (root)-[:LINK*0..]->(before),(after)-[:LINK*0..]->(root),(before)-[old:LINK]->(after)
WHERE root.name = 'ROOT' AND (before.value < 10 OR before = root) AND (10 < after.value OR after =
root)
CREATE UNIQUE (before)-[:LINK]->({ value:10 })-[:LINK]->(after)
DELETE old
35
nam e = 'ROOT'
LINK
value = 10
LINK
LINK
value = 25
Deleting a value, conversely, is done by finding the node with the value, and the two relationships
going in and out from it, and replacing the relationships with a new one.
MATCH (root)-[:LINK*0..]->(before),(before)-[delBefore:LINK]->(del)-[delAfter:LINK]->(after),
(after)-[:LINK*0..]->(root)
WHERE root.name = 'ROOT' AND del.value = 10
CREATE UNIQUE (before)-[:LINK]->(after)
DELETE del, delBefore, delAfter
nam e = 'ROOT'
LINK
LINK
value = 25
Deleting the last value node is what requires us to use CREATE UNIQUE when replacing the relationships.
Otherwise, we would end up with two relationships from the root node to itself, as both before and
after nodes are equal to the root node, meaning the pattern would match twice.
MATCH (root)-[:LINK*0..]->(before),(before)-[delBefore:LINK]->(del)-[delAfter:LINK]->(after),
(after)-[:LINK*0..]->(root)
WHERE root.name = 'ROOT' AND del.value = 25
CREATE UNIQUE (before)-[:LINK]->(after)
DELETE del, delBefore, delAfter
nam e = 'ROOT'
36
LINK
5.2.TV Shows
This example show how TV Shows with Seasons, Episodes, Characters, Actors, Users and Reviews
can be modeled in a graph database.
5.2.1.Data Model
Lets start out with an entity-relationship model of the domain at hand:
TV Show
has
Season
has
User
Episode
wrot e
has
Review
feat ured
Act or
played
Charact er
To implement this in Neo4j well use the following relationship types:

Relationship Type
Description
HAS_SEASON
Connects a show with its seasons.
HAS_EPISODE
Connects a season with its episodes.
FEATURED_CHARACTER
Connects an episode with its characters.
PLAYED_CHARACTER
Connects actors with characters. Note that an

actor can play multiple characters in an episode,
and that the same character can be played by
multiple actors as well.
HAS_REVIEW
Connects an episode with its reviews.
WROTE_REVIEW
Connects users with reviews they contributed.
5.2.2.Sample Data
Lets create some data and see how the domain plays out in practice:
CREATE (himym:TVShow { name: "How I Met Your Mother" })
37

CREATE (himym_s1:Season { name: "HIMYM Season 1" })
CREATE (himym_s1_e1:Episode { name: "Pilot" })
CREATE (ted:Character { name: "Ted Mosby" })
CREATE (joshRadnor:Actor { name: "Josh Radnor" })
CREATE UNIQUE (joshRadnor)-[:PLAYED_CHARACTER]->(ted)
CREATE UNIQUE (himym)-[:HAS_SEASON]->(himym_s1)
CREATE UNIQUE (himym_s1)-[:HAS_EPISODE]->(himym_s1_e1)
CREATE UNIQUE (himym_s1_e1)-[:FEATURED_CHARACTER]->(ted)
CREATE (himym_s1_e1_review1 { title: "Meet Me At The Bar In 15 Minutes & Suit Up",
content: "It was awesome" })
CREATE (wakenPayne:User { name: "WakenPayne" })
CREATE (wakenPayne)-[:WROTE_REVIEW]->(himym_s1_e1_review1)<-[:HAS_REVIEW]-(himym_s1_e1)
This is how the data looks in the database:

TVShow
nam e = 'How I Met Your Mot her'
HAS_SEASON
Season
nam e = 'HIMYM Season 1'
HAS_EPISODE
Act or
nam e = 'Josh Radnor'
PLAYED_CHARACTER
Episode
User
nam e = 'Pilot '
nam e = 'WakenPayne'
FEATURED_CHARACTER
Charact er
HAS_REVIEW
WROTE_REVIEW
t it le = 'Meet Me At The Bar In 15 Minut es & Suit Up'

cont ent = 'It was awesom e'
nam e = 'Ted Mosby'
Note that even though we could have modeled the reviews as relationships with title and content
properties on them, we made them nodes instead. We gain a lot of flexibility in this way, for example
if we want to connect comments to each review.
Now lets add more data:
MATCH (himym:TVShow { name: "How I Met Your Mother" }),(himym_s1:Season),
(himym_s1_e1:Episode { name: "Pilot" }),
(himym)-[:HAS_SEASON]->(himym_s1)-[:HAS_EPISODE]->(himym_s1_e1)
CREATE (marshall:Character { name: "Marshall Eriksen" })
CREATE (robin:Character { name: "Robin Scherbatsky" })
CREATE (barney:Character { name: "Barney Stinson" })
CREATE (lily:Character { name: "Lily Aldrin" })
CREATE (jasonSegel:Actor { name: "Jason Segel" })
CREATE (cobieSmulders:Actor { name: "Cobie Smulders" })
CREATE (neilPatrickHarris:Actor { name: "Neil Patrick Harris" })
CREATE (alysonHannigan:Actor { name: "Alyson Hannigan" })
CREATE UNIQUE (jasonSegel)-[:PLAYED_CHARACTER]->(marshall)
CREATE UNIQUE (cobieSmulders)-[:PLAYED_CHARACTER]->(robin)
CREATE UNIQUE (neilPatrickHarris)-[:PLAYED_CHARACTER]->(barney)
CREATE UNIQUE (alysonHannigan)-[:PLAYED_CHARACTER]->(lily)
38

CREATE UNIQUE (himym_s1_e1)-[:FEATURED_CHARACTER]->(marshall)
CREATE UNIQUE (himym_s1_e1)-[:FEATURED_CHARACTER]->(robin)
CREATE UNIQUE (himym_s1_e1)-[:FEATURED_CHARACTER]->(barney)
CREATE UNIQUE (himym_s1_e1)-[:FEATURED_CHARACTER]->(lily)
CREATE (himym_s1_e1_review2 { title: "What a great pilot for a show :)",
content: "The humour is great." })
CREATE (atlasredux:User { name: "atlasredux" })
CREATE (atlasredux)-[:WROTE_REVIEW]->(himym_s1_e1_review2)<-[:HAS_REVIEW]-(himym_s1_e1)
5.2.3.Information for a show

For a particular TV show, show all the seasons and all the episodes and all the reviews and all the cast
members from that show, that is all of the information connected to that TV show.
MATCH (tvShow:TVShow)-[:HAS_SEASON]->(season)-[:HAS_EPISODE]->(episode)
WHERE tvShow.name = "How I Met Your Mother"
RETURN season.name, episode.name
season.name
episode.name
"HIMYM Season 1"
"Pilot"
1 row
We could also grab the reviews if there are any by slightly tweaking the query:
MATCH (tvShow:TVShow)-[:HAS_SEASON]->(season)-[:HAS_EPISODE]->(episode)
WITH season, episode
OPTIONAL MATCH (episode)-[:HAS_REVIEW]->(review)
RETURN season.name, episode.name, review
season.name
episode.name
review
"HIMYM Season 1"
"Pilot"
Node[5]{title:"Meet Me At The
Bar In 15 Minutes & Suit Up",
content:"It was awesome"}
"HIMYM Season 1"
"Pilot"
Node[15]{title:"What a
great pilot for a show :)",
content:"The humour is great. "}
2 rows
Now lets list the characters featured in a show. Note that in this query we only put identifiers on the
nodes we actually use later on. The other nodes of the path pattern are designated by ().
MATCH (tvShow:TVShow)-[:HAS_SEASON]->()-[:HAS_EPISODE]->()-[:FEATURED_CHARACTER]->(character)
RETURN DISTINCT character.name
character.name
"Ted Mosby"
"Marshall Eriksen"
"Robin Scherbatsky"
"Barney Stinson"
5 rows
39

character.name
"Lily Aldrin"
5 rows
Now lets look at how to get all cast members of a show.
MATCH
(tvShow:TVShow)-[:HAS_SEASON]->()-[:HAS_EPISODE]->(episode)-[:FEATURED_CHARACTER]->()<-[:PLAYED_CHARACTER]-(actor)
RETURN DISTINCT actor.name
actor.name
"Josh Radnor"
"Jason Segel"
"Cobie Smulders"
"Neil Patrick Harris"
"Alyson Hannigan"
5 rows
5.2.4.Information for an actor

First lets add another TV show that Josh Radnor appeared in:
CREATE (er:TVShow { name: "ER" })
CREATE (er_s9:Season { name: "ER S7" })
CREATE (er_s9_e17:Episode { name: "Peter's Progress" })
CREATE (tedMosby:Character { name: "The Advocate " })
CREATE UNIQUE (er)-[:HAS_SEASON]->(er_s9)
CREATE UNIQUE (er_s9)-[:HAS_EPISODE]->(er_s9_e17)
WITH er_s9_e17
MATCH (actor:Actor),(episode:Episode)
WHERE actor.name = "Josh Radnor" AND episode.name = "Peter's Progress"
WITH actor, episode
CREATE (keith:Character { name: "Keith" })
CREATE UNIQUE (actor)-[:PLAYED_CHARACTER]->(keith)
CREATE UNIQUE (episode)-[:FEATURED_CHARACTER]->(keith)
And now well create a query to find the episodes that he has appeared in:
MATCH (actor:Actor)-[:PLAYED_CHARACTER]->(character)<-[:FEATURED_CHARACTER]-(episode)
WHERE actor.name = "Josh Radnor"
RETURN episode.name AS Episode, character.name AS Character
Episode
Character
"Pilot"
"Ted Mosby"
"Peter's Progress"
"Keith"
2 rows
Now lets go for a similar query, but add the season and show to it as well.
MATCH (actor:Actor)-[:PLAYED_CHARACTER]->(character)<-[:FEATURED_CHARACTER]-(episode),
(episode)<-[:HAS_EPISODE]-(season)<-[:HAS_SEASON]-(tvshow)
WHERE actor.name = "Josh Radnor"
RETURN tvshow.name AS Show, season.name AS Season, episode.name AS Episode,
40

character.name AS Character
Show
Season
Episode
Character
"How I Met Your Mother" "HIMYM Season 1"
"Pilot"
"Ted Mosby"
"ER"
"Peter's Progress"
"Keith"
"ER S7"
2 rows
41
5.3.ACL structures in graphs

This example gives a generic overview of an approach to handling Access Control Lists (ACLs) in
graphs, and a simplified example with concrete queries.
5.3.1.Generic approach
In many scenarios, an application needs to handle security on some form of managed objects. This
example describes one pattern to handle this through the use of a graph structure and traversers
that build a full permissions-structure for any managed object with exclude and include overriding
possibilities. This results in a dynamic construction of ACLs based on the position and context of the
managed object.
The result is a complex security scheme that can easily be implemented in a graph structure,
supporting permissions overriding, principal and content composition, without duplicating data
anywhere.
Technique
As seen in the example graph layout, there are some key concepts in this domain model:
The managed content (folders and files) that are connected by HAS_CHILD_CONTENT relationships
The Principal subtree pointing out principals that can act as ACL members, pointed out by the
PRINCIPAL relationships.
The aggregation of principals into groups, connected by the IS_MEMBER_OF relationship. One principal
(user or group) can be part of many groups at the same time.
The SECURITY relationships, connecting the content composite structure to the principal composite
structure, containing a addition/removal modifier property ("+RW").
42

Constructing the ACL
The calculation of the effective permissions (e.g. Read, Write, Execute) for a principal for any given
ACL-managed node (content) follows a number of rules that will be encoded into the permissionstraversal:
Top-down-Traversal
This approach will let you define a generic permission pattern on the root content, and then refine that
for specific sub-content nodes and specific principals.
1. Start at the content node in question traverse upwards to the content root node to determine the
path to it.
2. Start with a effective optimistic permissions list of "all permitted" (111 in a bit encoded
ReadWriteExecute case) or 000 if you like pessimistic security handling (everything is forbidden
unless explicitly allowed).
3. Beginning from the topmost content node, look for any SECURITY relationships on it.
4. If found, look if the principal in question is part of the end-principal of the SECURITY relationship.
5. If yes, add the "+" permission modifiers to the existing permission pattern, revoke the "-"
permission modifiers from the pattern.
6. If two principal nodes link to the same content node, first apply the more generic prinipals
modifiers.
7. Repeat the security modifier search all the way down to the target content node, thus overriding
more generic permissions with the set on nodes closer to the target node.
The same algorithm is applicable for the bottom-up approach, basically just traversing from the target
content node upwards and applying the security modifiers dynamically as the traverser goes up.
Example
Now, to get the resulting access rights for e.g. "user 1" on the "My File.pdf" in a Top-Down approach
on the model in the graph above would go like:
1. Traveling upward, we start with "Root folder", and set the permissions to 11 initially (only
considering Read, Write).
2. There are two SECURITY relationships to that folder. User 1 is contained in both of them, but "root"
is more generic, so apply it first then "All principals" +W +R 11.
3. "Home" has no SECURITY instructions, continue.
4. "user1 Home" has SECURITY. First apply "Regular Users" (-R -W) 00, Then "user 1" (+R +W) 11.
5. The target node "My File.pdf" has no SECURITY modifiers on it, so the effective permissions for "User
1" on "My File.pdf" are ReadWrite 11.
5.3.2.Read-permission example
In this example, we are going to examine a tree structure of directories and files. Also, there are
users that own files and roles that can be assigned to users. Roles can have permissions on directory or
files structures (here we model only canRead, as opposed to full rwx Unix permissions) and be nested.
A more thorough example of modeling ACL structures can be found at How to Build Role-Based
Access Control in SQL <http://www.xaprb.com/blog/2006/08/16/how-to-build-role-based-accesscontrol-in-sql/>.
43

Node[ 19]
'nam e' = 'Root '
has
Node[ 21]
'nam e' = 'Role'
subRole
subRole
has
Node[ 7]
Node[ 22]
'nam e' = 'User'
'nam e' = 'SUDOers'
m em ber
m em ber
m em ber
m em ber
canRead
Node[ 13]
Node[ 14]
Node[ 8]
Node[ 9]
Node[ 18]
'nam e' = 'User2'
'nam e' = 'User1'
'nam e' = 'Adm in2'
'nam e' = 'Adm in1'
'nam e' = 'FileRoot '

cont ains
cont ains
Node[ 10]
Node[ 12]
'nam e' = 'Hom e'
'nam e' = 'et c'
owns
cont ains
cont ains
cont ains
Node[ 20]
Node[ 15]
Node[ 11]
'nam e' = 'Hom eU1'
'nam e' = 'Hom eU2'
'nam e' = 'init .d'
owns
leaf
cont ains
Node[ 17]
Node[ 23]
'nam e' = 'File1'
'nam e' = 'Deskt op'

leaf
Node[ 16]
'nam e' = 'File2'
Find all files in the directory structure

In order to find all files contained in this structure, we need a variable length query that follows all
contains relationships and retrieves the nodes at the other end of the leaf relationships.
MATCH ({ name: 'FileRoot' })-[:contains*0..]->(parentDir)-[:leaf]->(file)
RETURN file
resulting in:
file
Node[10]{name:"File1"}
2 rows
What files are owned by whom?
If we introduce the concept of ownership on files, we then can ask for the owners of the files we
find connected via owns relationships to file nodes.
MATCH ({ name: 'FileRoot' })-[:contains*0..]->()-[:leaf]->(file)<-[:owns]-(user)
RETURN file, user
Returning the owners of all files below the FileRoot node.

file
user
Node[7]{name:"User1"}
Node[6]{name:"User2"}
2 rows
44

Who has access to a File?
If we now want to check what users have read access to all Files, and define our ACL as
The root directory has no access granted.
Any user having a role that has been granted canRead access to one of the parent folders of a File has
read access.
In order to find users that can read any part of the parent folder hierarchy above the files, Cypher
provides optional variable length path.
MATCH (file)<-[:leaf]-()<-[:contains*0..]-(dir)
OPTIONAL MATCH (dir)<-[:canRead]-(role)-[:member]->(readUser)
WHERE file.name =~ 'File.*'
RETURN file.name, dir.name, role.name, readUser.name
This will return the file, and the directory where the user has the canRead permission along with the
user and their role.
file.name
dir.name
role.name
readUser.name
"File1"
"HomeU1"
<null>
<null>
"File1"
"Home"
<null>
<null>
"File1"
"FileRoot"
"SUDOers"
"Admin1"
"File1"
"FileRoot"
"SUDOers"
"Admin2"
"File2"
"Desktop"
<null>
<null>
"File2"
"HomeU2"
<null>
<null>
"File2"
"Home"
<null>
<null>
"File2"
"FileRoot"
"SUDOers"
"Admin1"
"File2"
"FileRoot"
"SUDOers"
"Admin2"
9 rows
The results listed above contain null for optional path segments, which can be mitigated by either
asking several queries or returning just the really needed values.
45
5.4.Hyperedges
Imagine a user being part of different groups. A group can have different roles, and a user can be part
of different groups. He also can have different roles in different groups apart from the membership.
The association of a User, a Group and a Role can be referred to as a HyperEdge. However, it can be
easily modeled in a property graph as a node that captures this n-ary relationship, as depicted below in
the U1G2R1 node.
Figure5.1.Graph
nam e = 'User1'
hasRoleInGroup
nam e = 'U1G2R1'
hasRoleInGroup
in
in
nam e = 'U1G1R2'
hasGroup
hasRole
hasGroup
nam e = 'Group2'
canHave
nam e = 'Group1'
isA
canHave
nam e = 'Role1'
canHave
isA
nam e = 'Group'
isA
hasRole
canHave
nam e = 'Role2'
isA
nam e = 'Role'
5.4.1.Find Groups
To find out in what roles a user is for a particular groups (here Group2), the following query can
traverse this HyperEdge node and provide answers.
Query.
MATCH ({ name: 'User1' })-[:hasRoleInGroup]->(hyperEdge)-[:hasGroup]->({ name: 'Group2' }),
(hyperEdge)-[:hasRole]->(role)
RETURN role.name
The role of User1 is returned:
46

Result
role.name
"Role1"
1 row
5.4.2.Find all groups and roles for a user

Here, find all groups and the roles a user has, sorted by the name of the role.
Query.
MATCH ({ name: 'User1' })-[:hasRoleInGroup]->(hyperEdge)-[:hasGroup]->(group),
(hyperEdge)-[:hasRole]->(role)
RETURN role.name, group.name
ORDER BY role.name ASC
The groups and roles of User1 are returned:

Result
role.name
group.name
"Role1"
"Group2"
"Role2"
"Group1"
2 rows
5.4.3.Find common groups based on shared roles

Assume a more complicated graph:
1.
2.
3.
4.
5.
Two user nodes User1, User2.

User1 is in Group1, Group2, Group3.
User1 has Role1, Role2 in Group1; Role2, Role3 in Group2; Role3, Role4 in Group3 (hyper edges).
User2 is in Group1, Group2, Group3.
User2 has Role2, Role5 in Group1; Role3, Role4 in Group2; Role5, Role6 in Group3 (hyper edges).
The graph for this looks like the following (nodes like U1G2R23 representing the HyperEdges):
Figure5.2.Graph
nam e = 'User2'
hasRoleInGroup
nam e = 'U2G2R34'
hasGroup
nam e = 'Group2'
hasGroup
hasRole
hasRole
nam e = 'Role3'
hasRoleInGroup
nam e = 'U1G2R23'
hasRole
hasRole
nam e = 'Role4'
hasRoleInGroup
nam e = 'U1G3R34'
hasRole
nam e = 'User1'
hasGroup
nam e = 'Group3'
hasRoleInGroup
hasRoleInGroup
hasRoleInGroup
nam e = 'U2G3R56'
hasGroup
hasRole
nam e = 'Role6'
nam e = 'U2G1R25'
hasRole
hasRole
nam e = 'Role5'
hasRole
nam e = 'U1G1R12'
hasRole
nam e = 'Role2'
hasRole
hasGroup
hasGroup
nam e = 'Group1'
hasRole
nam e = 'Role1'
To return Group1 and Group2 as User1 and User2 share at least one common role in these two groups, the
query looks like this:
Query.
MATCH (u1)-[:hasRoleInGroup]->(hyperEdge1)-[:hasGroup]->(group),(hyperEdge1)-[:hasRole]->(role),
(u2)-[:hasRoleInGroup]->(hyperEdge2)-[:hasGroup]->(group),(hyperEdge2)-[:hasRole]->(role)
WHERE u1.name = 'User1' AND u2.name = 'User2'
RETURN group.name, count(role)
47

ORDER BY group.name ASC
The groups where User1 and User2 share at least one common role:
Result
group.name
count(role)
"Group1"
"Group2"
2 rows
48
5.5.Basic friend finding based on social neighborhood

Imagine an example graph like the following one:
Figure5.3.Graph
nam e = 'Joe'
knows
knows
nam e = 'Sara'
knows
knows
nam e = 'Bill'
knows
nam e = 'Derrick'
knows
nam e = 'Jill'
knows
nam e = 'Ian'
To find out the friends of Joes friends that are not already his friends, the query looks like this:
Query.
MATCH (joe { name: 'Joe' })-[:knows*2..2]-(friend_of_friend)
WHERE NOT (joe)-[:knows]-(friend_of_friend)
RETURN friend_of_friend.name, COUNT(*)
ORDER BY COUNT(*) DESC , friend_of_friend.name
This returns a list of friends-of-friends ordered by the number of connections to them, and secondly by
their name.
Result
friend_of_friend.name
COUNT(*)
"Ian"
"Derrick"
"Jill"
3 rows
49
5.6.Co-favorited places
Figure5.4.Graph
nam e = 'Joe'
favorit e
nam e = 'SaunaX'
favorit e
nam e = 'CoffeeShop1'
nam e = 'Jill'
favorit e
favorit e
favorit e
t agged
nam e = 'MelsPlace'
t agged
t agged
t agged
nam e = 'Cosy'
favorit e
t agged
nam e = 'CoffeShop2'
t agged
nam e = 'Cool'
5.6.1.Co-favorited places users who like x also like y

Find places that people also like who favorite this place:
Determine who has favorited place x.
What else have they favorited that is not place x.
Query.
MATCH (place)<-[:favorite]-(person)-[:favorite]->(stuff)
WHERE place.name = 'CoffeeShop1'
RETURN stuff.name, count(*)
ORDER BY count(*) DESC , stuff.name
The list of places that are favorited by people that favorited the start place.
Result
stuff.name
count(*)
"MelsPlace"
"CoffeShop2"
"SaunaX"
3 rows
5.6.2.Co-Tagged places places related through tags

Find places that are tagged with the same tags:
Determine the tags for place x.
What else is tagged the same as x that is not x.
Query.
MATCH (place)-[:tagged]->(tag)<-[:tagged]-(otherPlace)
WHERE place.name = 'CoffeeShop1'
RETURN otherPlace.name, collect(tag.name)
ORDER BY length(collect(tag.name)) DESC , otherPlace.name
This query returns other places than CoffeeShop1 which share the same tags; they are ranked by the
number of tags.
50

Result
otherPlace.name
collect(tag.name)
"MelsPlace"
["Cool", "Cosy"]
"CoffeeShop2"
["Cool"]
"CoffeeShop3"
["Cosy"]
3 rows
51
5.7.Find people based on similar favorites

Figure5.5.Graph
nam e = 'Joe'
friend
nam e = 'Jill'
favorit e
favorit e
nam e = 'Sara'
favorit e
favorit e
nam e = 'Bikes'
nam e = 'Derrick'
favorit e
favorit e
favorit e
nam e = 'Cat s'
To find out the possible new friends based on them liking similar things as the asking person, use a
query like this:
Query.
MATCH (me { name: 'Joe' })-[:favorite]->(stuff)<-[:favorite]-(person)
WHERE NOT (me)-[:friend]-(person)
RETURN person.name, count(stuff)
ORDER BY count(stuff) DESC
The list of possible friends ranked by them liking similar stuff that are not yet friends is returned.
Result
person.name
count(stuff)
"Derrick"
"Jill"
2 rows
52
5.8.Find people based on mutual friends and groups

Figure5.6.Graph
Node[ 3] nam e = 'Jill'
Node[ 4] nam e = 'Joe'
knows
m em ber_of_group
Node[ 2] nam e = 'Bob'
knows
Node[ 0] nam e = 'Bill'
m em ber_of_group
m em ber_of_group
m em ber_of_group
Node[ 1] nam e = 'Group1'
In this scenario, the problem is to determine mutual friends and groups, if any, between persons. If no
mutual groups or friends are found, there should be a 0 returned.
Query.
MATCH (me { name: 'Joe' }),(other)
WHERE other.name IN ['Jill', 'Bob']
OPTIONAL MATCH pGroups=(me)-[:member_of_group]->(mg)<-[:member_of_group]-(other)
OPTIONAL MATCH pMutualFriends=(me)-[:knows]->(mf)<-[:knows]-(other)
RETURN other.name AS name, count(DISTINCT pGroups) AS mutualGroups,
count(DISTINCT pMutualFriends) AS mutualFriends
ORDER BY mutualFriends DESC
The question we are asking is how many unique paths are there between me and Jill, the paths
being common group memberships, and common friends. If the paths are mandatory, no results will
be returned if me and Bob lack any common friends, and we dont want that. To make a path optional,
you have to make at least one of its relationships optional. That makes the whole path optional.
Result
name
mutualGroups
mutualFriends
"Jill"
"Bob"
2 rows
53
5.9.Find friends based on similar tagging

Figure5.7.Graph
nam e = 'Joe'
favorit e
nam e = 'Cat s'
favorit e
nam e = 'Sara'
favorit e
nam e = 'Horses'
t agged
favorit e
favorit e
nam e = 'Surfing'
t agged
t agged
nam e = 'Anim als'
nam e = 'Derrick'
favorit e
favorit e
nam e = 'Bikes'
t agged
nam e = 'Hobby'
To find people similar to me based on the taggings of their favorited items, one approach could be:
Determine the tags associated with what I favorite.

What else is tagged with those tags?
Who favorites items tagged with the same tags?
Sort the result by how many of the same things these people like.
Query.
MATCH
(me)-[:favorite]->(myFavorites)-[:tagged]->(tag)<-[:tagged]-(theirFavorites)<-[:favorite]-(people)
WHERE me.name = 'Joe' AND NOT me=people
RETURN people.name AS name, count(*) AS similar_favs
ORDER BY similar_favs DESC
The query returns the list of possible friends ranked by them liking similar stuff that are not yet
friends.
Result
name
similar_favs
"Sara"
"Derrick"
2 rows
54
5.10.Multirelational (social) graphs

Figure5.8.Graph
nam e = 'cars'
LIKES
nam e = 'Ben'
LIKES
nam e = 'Maria'
FOLLOWS
nam e = 'Sara'
LIKES
nam e = 'cat s'
FOLLOWS
LOVES
FOLLOWS
LOVES
LIKES
FOLLOWS
FOLLOWS
nam e = 'Joe'
LIKES
nam e = 'bikes'
LIKES
nam e = 'nat ure'
This example shows a multi-relational network between persons and things they like. A multirelational graph is a graph with more than one kind of relationship between nodes.
Query.
MATCH (me { name: 'Joe' })-[r1:FOLLOWS|:LOVES]->(other)-[r2]->(me)
WHERE type(r1)=type(r2)
RETURN other.name, type(r1)
The query returns people that FOLLOWS or LOVES Joe back.

Result
other.name
type(r1)
"Sara"
"FOLLOWS"
"Maria"
"FOLLOWS"
"Maria"
"LOVES"
3 rows
55
5.11.Implementing newsfeeds in a graph

nam e = 'Bob'
STATUS
nam e = 'bob_s1'
t ext = 'bobs st at us1'
dat e = 1
NEXT
nam e = 'bob_s2'
dat e = 4
FRIEND
st at us = 'CONFIRMED'
FRIEND
st at us = 'CONFIRMED'
nam e = 'Alice'
STATUS
FRIEND
st at us = 'PENDING'
nam e = 'alice_s1'
t ext = 'Alices st at us1'
dat e = 2
nam e = 'Joe'
NEXT
nam e = 'alice_s2'
t ext = 'Alices st at us2'
dat e = 5
STATUS
nam e = 'joe_s1'
t ext = 'Joe st at us1'
dat e = 3
NEXT
nam e = 'joe_s2'
t ext = 'Joe st at us2'
dat e = 6
Implementation of newsfeed or timeline feature is a frequent requirement for social applications. The
following exmaples are inspired by Newsfeed feature powered by Neo4j Graph Database <http://
techfin.in/2012/10/newsfeed-feature-powered-by-neo4j-graph-database/>. The query asked here is:
Starting at me, retrieve the time-ordered status feed of the status updates of me and and all friends that
are connected via a CONFIRMED FRIEND relationship to me.
Query.
MATCH (me { name: 'Joe' })-[rels:FRIEND*0..1]-(myfriend)
WHERE ALL (r IN rels WHERE r.status = 'CONFIRMED')
WITH myfriend
MATCH (myfriend)-[:STATUS]-(latestupdate)-[:NEXT*0..1]-(statusupdates)
RETURN myfriend.name AS name, statusupdates.date AS date, statusupdates.text AS text
ORDER BY statusupdates.date DESC LIMIT 3
To understand the strategy, lets divide the query into five steps:
1. First Get the list of all my friends (along with me) through FRIEND relationship (MATCH (me {name:
'Joe'})-[rels:FRIEND*0..1]-(myfriend)). Also, the WHERE predicate can be added to check whether
the friend request is pending or confirmed.
56

2. Get the latest status update of my friends through Status relationship (MATCH myfriend-[:STATUS]latestupdate).
3. Get subsequent status updates (along with the latest one) of my friends through NEXT relationships
(MATCH (myfriend)-[:STATUS]-(latestupdate)-[:NEXT*0..1]-(statusupdates)) which will give you the
latest and one additional statusupdate; adjust 0..1 to whatever suits your case.
4. Sort the status updates by posted date (ORDER BY statusupdates.date DESC).
5. LIMIT the number of updates you need in every query (LIMIT 3).
Result
name
date
text
"Joe"
"Joe status2"
"Bob"
"bobs status2"
"Joe"
"Joe status1"
3 rows
Here, the example shows how to add a new status update into the existing data for a user.
Query.
MATCH (me)
WHERE me.name='Bob'
OPTIONAL MATCH (me)-[r:STATUS]-(secondlatestupdate)
DELETE r
CREATE (me)-[:STATUS]->(latest_update { text:'Status',date:123 })
WITH latest_update, collect(secondlatestupdate) AS seconds
FOREACH (x IN seconds | CREATE latest_update-[:NEXT]->x)
RETURN latest_update.text AS new_status
Dividing the query into steps, this query resembles adding new item in middle of a doubly linked list:
1. Get the latest update (if it exists) of the user through the STATUS relationship (OPTIONAL MATCH (me)[r:STATUS]-(secondlatestupdate)).
2. Delete the STATUS relationship between user and secondlatestupdate (if it exists), as this would
become the second latest update now and only the latest update would be added through a STATUS
relationship; all earlier updates would be connected to their subsequent updates through a NEXT
relationship. (DELETE r).
3. Now, create the new statusupdate node (with text and date as properties) and connect
this with the user through a STATUS relationship (CREATE me-[:STATUS]->(latest_update
{ text:'Status',date:123 })).
4. Pipe over statusupdate or an empty collection to the next query part (WITH latest_update,
collect(secondlatestupdate) AS seconds).
5. Now, create a NEXT relationship between the latest status update and the second latest status update
(if it exists) (FOREACH(x in seconds | CREATE latest_update-[:NEXT]->x)).
57

Result
new_status
"Status"
1 row
Nodes created: 1
Relationships created: 2
Properties set: 2
Relationships deleted: 1
Node[ 0] nam e = 'Bob'
STATUS
Node[ 1] nam e = 'bob_s1'

dat e = 1
NEXT
Node[ 2] nam e = 'bob_s2'

dat e = 4
58
5.12.Boosting recommendation results

Figure5.9.Graph
nam e = 'Clark Kent '
KNOWS
weight = 4
nam e = 'Lois Lane'
KNOWS
weight = 4
nam e = 'Jim m y Olsen'
KNOWS
weight = 4
WORKS_AT
weight = 2
act ivit y = 56
nam e = 'Anderson Cooper'
KNOWS
weight = 4
WORKS_AT
weight = 2
act ivit y = 10
nam e = 'Perry Whit e'
WORKS_AT
weight = 2
act ivit y = 2
WORKS_AT
weight = 2
act ivit y = 3
nam e = 'CNN'
WORKS_AT
weight = 2
act ivit y = 45
KNOWS
weight = 4
WORKS_AT
weight = 2
act ivit y = 6
nam e = 'Daily Planet '
This query finds the recommended friends for the origin that are working at the same place as the
origin, or know a person that the origin knows, also, the origin should not already know the target.
This recommendation is weighted for the weight of the relationship r2, and boosted with a factor of 2,
if there is an activity-property on that relationship
Query.
MATCH (origin)-[r1:KNOWS|WORKS_AT]-(c)-[r2:KNOWS|WORKS_AT]-(candidate)
WHERE origin.name = "Clark Kent" AND type(r1)=type(r2) AND NOT (origin)-[:KNOWS]-(candidate)
RETURN origin.name AS origin, candidate.name AS candidate, SUM(ROUND(r2.weight
+(COALESCE(r2.activity,
0)* 2))) AS boost
ORDER BY boost DESC LIMIT 10
This returns the recommended friends for the origin nodes and their recommendation score.
Result
origin
candidate
boost
"Clark Kent"
"Perry White"
22
"Clark Kent"
"Anderson Cooper"
2 rows
59
5.13.Calculating the clustering coefficient of a network

Figure5.10.Graph
nam e = 'st art node'
KNOWS
KNOWS
KNOWS
KNOWS
KNOWS
KNOWS
KNOWS
In this example, adapted from Niko Gamulins blog post on Neo4j for Social Network Analysis
<http://mypetprojects.blogspot.se/2012/06/social-network-analysis-with-neo4j.html>, the graph in
question is showing the 2-hop relationships of a sample person as nodes with KNOWS relationships.
The clustering coefficient <http://en.wikipedia.org/wiki/Clustering_coefficient> of a selected node
is defined as the probability that two randomly selected neighbors are connected to each other. With
the number of neighbors as n and the number of mutual connections between the neighbors r the
calculation is:
The number of possible connections between two neighbors is n!/(2!(n-2)!) = 4!/(2!(4-2)!) = 24/4 =
6, where n is the number of neighbors n = 4 and the actual number r of connections is 1. Therefore the
clustering coefficient of node 1 is 1/6.
and r are quite simple to retrieve via the following query:
Query.
MATCH (a { name: "startnode" })--(b)
WITH a, count(DISTINCT b) AS n
MATCH (a)--()-[r]-()--(a)
RETURN n, count(DISTINCT r) AS r
This returns n and r for the above calculations.

Result
n
1 row
60
5.14.Pretty graphs
This section is showing how to create some of the named pretty graphs on Wikipedia <http://
en.wikipedia.org/wiki/Gallery_of_named_graphs>.
5.14.1.Star graph
The graph is created by first creating a center node, and then once per element in the range, creates a
leaf node and connects it to the center.
Query.
CREATE (center)
FOREACH (x IN range(1,6)| CREATE (leaf),(center)-[:X]->(leaf))
RETURN id(center) AS id;
The query returns the id of the center node.

Result
id
7
1 row
Nodes created: 7
Figure5.11.Graph
X
X
X
X
X
X
5.14.2.Wheel graph
This graph is created in a number of steps:
Create a center node.

Once per element in the range, create a leaf and connect it to the center.
Connect neighboring leafs.
Find the minimum and maximum leaf and connect these.
61

Return the id of the center node.
Query.
CREATE (center)
FOREACH (x IN range(1,6)| CREATE (leaf { count:x }),(center)-[:X]->(leaf))
WITH center
MATCH (large_leaf)<--(center)-->(small_leaf)
WHERE large_leaf.count = small_leaf.count + 1
CREATE (small_leaf)-[:X]->(large_leaf)
WITH center, min(small_leaf.count) AS min, max(large_leaf.count) AS max
MATCH (first_leaf)<--(center)-->(last_leaf)
WHERE first_leaf.count = min AND last_leaf.count = max
CREATE (last_leaf)-[:X]->(first_leaf)
RETURN id(center) AS id
The query returns the id of the center node.

Result
id
7
1 row
Nodes created: 7
Properties set: 6
Figure5.12.Graph
count = 4
X
count = 3
X
X
X
count = 5
X
X
count = 2
X
X
count = 6
X
count = 1
5.14.3.Complete graph
For this graph, a root node is created, and used to hang a number of nodes from. Then, two nodes are
selected, hanging from the center, with the requirement that the id of the first is less than the id of the
next. This is to prevent double relationships and self relationships. Using said match, relationships
between all these nodes are created. Lastly, the center node and all relationships connected to it are
removed.
Query.
CREATE (center)
62

FOREACH (x IN range(1,6)| CREATE (leaf { count : x }),(center)-[:X]->(leaf))
WITH center
MATCH (leaf1)<--(center)-->(leaf2)
WHERE id(leaf1)<id(leaf2)
CREATE (leaf1)-[:X]->(leaf2)
WITH center
MATCH (center)-[r]->()
DELETE center,r;
Nothing is returned by this query.

Result
(empty result)
Nodes created: 7
Properties set: 6
Nodes deleted: 1
Figure5.13.Graph
count = 4
count = 5
X
X
count = 3
X
X
count = 2
count = 6
count = 1
5.14.4.Friendship graph
This query first creates a center node, and then once per element in the range, creates a cycle graph
and connects it to the center
Query.
CREATE (center)
FOREACH (x IN range(1,3)| CREATE (leaf1),(leaf2),(center)-[:X]->(leaf1),(center)-[:X]->(leaf2),
(leaf1)-[:X]->(leaf2))
RETURN ID(center) AS id
The id of the center node is returned by the query.

63

Result
id
7
1 row
Nodes created: 7
Figure5.14.Graph
X
X
X
X
X
X
X
X
64
5.15.A multilevel indexing structure (path tree)

In this example, a multi-level tree structure is used to index event nodes (here Event1, Event2 and
Event3, in this case with a YEAR-MONTH-DAY granularity, making this a timeline indexing
structure. However, this approach should work for a wide range of multi-level ranges.
The structure follows a couple of rules:
Events can be indexed multiple times by connecting the indexing structure leafs with the events via
a VALUE relationship.
The querying is done in a path-range fashion. That is, the start- and end path from the indexing root
to the start and end leafs in the tree are calculated
Using Cypher, the queries following different strategies can be expressed as path sections and put
together using one single query.
The graph below depicts a structure with 3 Events being attached to an index structure at different
leafs.
Figure5.15.Graph
Root
2010
2011
Year 2010
Year 2011
12
01
Mont h 12
Mont h 01
31
Day 31
VALUE
01
NEXT
VALUE
Event 1
Day 01
02
NEXT
VALUE
03
Day 02
NEXT
Day 03
VALUE
Event 2
Event 3
5.15.1.Return zero range

Here, only the events indexed under one leaf (2010-12-31) are returned. The query only needs one
path segment rootPath (color Green) through the index.
65

Figure5.16.Graph
Root
2010
2011
Year 2010
Year 2011
12
01
Mont h 12
Mont h 01
31
Day 31
VALUE
Event 1
01
NEXT
Day 01
VALUE
02
NEXT
03
Day 02
NEXT
VALUE
Day 03
VALUE
Event 2
Event 3
Query.
MATCH rootPath=(root)-[:`2010`]->()-[:`12`]->()-[:`31`]->(leaf),(leaf)-[:VALUE]->(event)
WHERE root.name = 'Root'
RETURN event.name
ORDER BY event.name ASC
Returning all events on the date 2010-12-31, in this case Event1 and Event2
Result
event.name
"Event1"
"Event2"
2 rows
5.15.2.Return the full range

In this case, the range goes from the first to the last leaf of the index tree. Here, startPath (color
Greenyellow) and endPath (color Green) span up the range, valuePath (color Blue) is then connecting the
leafs, and the values can be read from the middle node, hanging off the values (color Red) path.
66

Figure5.17.Graph
Root
2010
2011
Year 2010
Year 2011
12
01
Mont h 12
Mont h 01
31
Day 31
VALUE
Event 1
01
NEXT
VALUE
Day 01
02
NEXT
03
Day 02
NEXT
VALUE
Day 03
VALUE
Event 2
Event 3
Query.
MATCH startPath=(root)-[:`2010`]->()-[:`12`]->()-[:`31`]->(startLeaf),
endPath=(root)-[:`2011`]->()-[:`01`]->()-[:`03`]->(endLeaf),
valuePath=(startLeaf)-[:NEXT*0..]->(middle)-[:NEXT*0..]->(endLeaf),
vals=(middle)-[:VALUE]->(event)
RETURN event.name
Returning all events between 2010-12-31 and 2011-01-03, in this case all events.
Result
event.name
"Event1"
"Event2"
"Event2"
"Event3"
4 rows
67
5.15.3.Return partly shared path ranges

Here, the query range results in partly shared paths when querying the index, making the introduction
of and common path segment commonPath (color Black) necessary, before spanning up startPath (color
Greenyellow) and endPath (color Darkgreen) . After that, valuePath (color Blue) connects the leafs and the
indexed values are returned off values (color Red) path.
Figure5.18.Graph
Root
2010
2011
Year 2010
Year 2011
12
01
Mont h 12
Mont h 01
31
Day 31
VALUE
Event 1
01
NEXT
VALUE
Day 01
02
NEXT
03
Day 02
NEXT
VALUE
Day 03
VALUE
Event 2
Event 3
Query.
MATCH commonPath=(root)-[:`2011`]->()-[:`01`]->(commonRootEnd),
startPath=(commonRootEnd)-[:`01`]->(startLeaf), endPath=(commonRootEnd)-[:`03`]->(endLeaf),
valuePath=(startLeaf)-[:NEXT*0..]->(middle)-[:NEXT*0..]->(endLeaf),
vals=(middle)-[:VALUE]->(event)
RETURN event.name
Returning all events between 2011-01-01 and 2011-01-03, in this case Event2 and Event3.
Result
event.name
"Event2"
2 rows
68

event.name
"Event3"
2 rows
69
5.16.Complex similarity computations

5.16.1.Calculate similarities by complex calculations
Here, a similarity between two players in a game is calculated by the number of times they have eaten
the same food.
Query.
MATCH (me { name: 'me' })-[r1:ATE]->(food)<-[r2:ATE]-(you)
WITH me,count(DISTINCT r1) AS H1,count(DISTINCT r2) AS H2,you
MATCH (me)-[r1:ATE]->(food)<-[r2:ATE]-(you)
RETURN sum((1-ABS(r1.times/H1-r2.times/H2))*(r1.times+r2.times)/(H1+H2)) AS similarity
The two players and their similarity measure.

Result
similarity
-30. 0
1 row
Figure5.19.Graph
nam e = 'm e'
nam e = 'you'
ATE
t im es = 10
nam e = 'm eat '
70
ATE
t im es = 5
5.17.The Graphity activity stream model

5.17.1.Find Activity Streams in a network without scaling penalty
This is an approach for scaling the retrieval of activity streams in a friend graph put forward by Rene
Pickard as Graphity <http://www.rene-pickhardt.de/graphity-an-efficient-graph-model-for-retrievingthe-top-k-news-feeds-for-users-in-social-networks/>. In short, a linked list is created for every persons
friends in the order that the last activities of these friends have occured. When new activities occur for
a friend, all the ordered friend lists that this friend is part of are reordered, transferring computing load
to the time of new event updates instead of activity stream reads.
Tip
This approach of course makes excessive use of relationship types. This needs to be
taken into consideration when designing a production system with this approach. See
Section15.5, Capacity for the maximum number of relationship types.
To find the activity stream for a person, just follow the linked list of the friend list, and retrieve the
needed amount of activities form the respective activity list of the friends.
Query.
MATCH p=(me { name: 'Jane' })-[:jane_knows*]->(friend),(friend)-[:has]->(status)
RETURN me.name, friend.name, status.name, length(p)
ORDER BY length(p)
The returns the activity stream for Jane.

Result
me.name
friend.name
status.name
length(p)
"Jane"
"Bill"
"Bill_s1"
"Jane"
"Joe"
"Joe_s1"
"Jane"
"Bob"
"Bob_s1"
3 rows
71

Figure5.20.Graph
nam e = 'Jane'
jane_knows
nam e = 'Bill'
has
jane_knows
nam e = 'Bill_s1'
nam e = 'Joe'
next
nam e = 'Bill_s2'
has
nam e = 'Joe_s1'
jane_knows
bob_knows
nam e = 'Bob'
next
has
nam e = 'Joe_s2'
nam e = 'Bob_s1'
bob_knows
nam e = 'Ted'
has
nam e = 'Ted_s1'
next
nam e = 'Ted_s2'
72
5.18.User roles in graphs
This is an example showing a hierarchy of roles. Whats interesting is that a tree is not sufficient for
storing this kind of structure, as elaborated below.
This is an implementation of an example found in the article A Model to Represent Directed Acyclic
Graphs (DAG) on SQL Databases <http://www.codeproject.com/Articles/22824/A-Model-toRepresent-Directed-Acyclic-Graphs-DAG-o> by Kemal Erdogan <http://www.codeproject.com/
script/Articles/MemberArticles.aspx?amid=274518>. The article discusses how to store directed
acyclic graphs <http://en.wikipedia.org/wiki/Directed_acyclic_graph> (DAGs) in SQL based DBs.
DAGs are almost trees, but with a twist: it may be possible to reach the same node through different
paths. Trees are restricted from this possibility, which makes them much easier to handle. In our case
it is Ali and Engin, as they are both admins and users and thus reachable through these group
nodes. Reality often looks this way and cant be captured by tree structures.
In the article an SQL Stored Procedure solution is provided. The main idea, that also have some
support from scientists, is to pre-calculate all possible (transitive) paths. Pros and cons of this
approach:
decent performance on read

low performance on insert
wastes lots of space
relies on stored procedures
In Neo4j storing the roles is trivial. In this case we use PART_OF (green edges) relationships to model
the group hierarchy and MEMBER_OF (blue edges) to model membership in groups. We also connect the
top level groups to the reference node by ROOT relationships. This gives us a useful partitioning of the
graph. Neo4j has no predefined relationship types, you are free to create any relationship types and
give them the semantics you want.
Lets now have a look at how to retrieve information from the graph. The the queries are done using
Cypher, the Java code is using the Neo4j Traversal API (see Section33.2, Traversal Framework Java
API, which is part of PartVIII, Advanced Usage).
73
5.18.1.Get the admins

In Cypher, we could get the admins like this:
MATCH ({ name: 'Admins' })<-[:PART_OF*0..]-(group)<-[:MEMBER_OF]-(user)
RETURN user.name, group.name
resulting in:
user.name
group.name
"Demet"
"HelpDesk"
"Ali"
"Admins"
"Engin"
"HelpDesk"
3 rows
And heres the code when using the Java Traversal API:
Node admins = getNodeByName( "Admins" );
TraversalDescription traversalDescription = db.traversalDescription()
.breadthFirst()
.evaluator( Evaluators.excludeStartPosition() )
.relationships( RoleRels.PART_OF, Direction.INCOMING )
.relationships( RoleRels.MEMBER_OF, Direction.INCOMING );
Traverser traverser = traversalDescription.traverse( admins );
resulting in the output

Found:
Found:
Found:
Found:
HelpDesk at depth: 0
Ali at depth: 0
Engin at depth: 1
Demet at depth: 1
The result is collected from the traverser using this code:

String output = "";
for ( Path path : traverser )
{
Node node = path.endNode();
output += "Found: " + node.getProperty( NAME ) + " at depth: "
+ ( path.length() - 1 ) + "\n";
}
5.18.2.Get the group memberships of a user

In Cypher:
MATCH ({ name: 'Jale' })-[:MEMBER_OF]->()-[:PART_OF*0..]->(group)
RETURN group.name
group.name
"ABCTechnicians"
"Technicians"
"Users"
3 rows
Using the Neo4j Java Traversal API, this query looks like:
Node jale = getNodeByName( "Jale" );
74

traversalDescription = db.traversalDescription()
.depthFirst()
.relationships( RoleRels.MEMBER_OF, Direction.OUTGOING )
.relationships( RoleRels.PART_OF, Direction.OUTGOING );
traverser = traversalDescription.traverse( jale );
resulting in:
Found: ABCTechnicians at depth: 0
Found: Technicians at depth: 1
Found: Users at depth: 2
5.18.3.Get all groups

In Cypher:
MATCH ({ name: 'Reference_Node' })<-[:ROOT]->()<-[:PART_OF*0..]-(group)
RETURN group.name
group.name
"Users"
"ABCTechnicians"
"Managers"
"HelpDesk"
"Technicians"
"Admins"
6 rows
In Java:
Node referenceNode = getNodeByName( "Reference_Node") ;
.breadthFirst()
.relationships( RoleRels.ROOT, Direction.INCOMING )
.relationships( RoleRels.PART_OF, Direction.INCOMING );
traverser = traversalDescription.traverse( referenceNode );
resulting in:
Found:
Found:
Found:
Found:
Found:
Found:
Admins at depth: 0
Users at depth: 0
HelpDesk at depth: 1
Managers at depth: 1
Technicians at depth: 1
ABCTechnicians at depth: 2
5.18.4.Get all members of all groups

Now, lets try to find all users in the system being part of any group.
In Cypher, this looks like:
MATCH ({ name: 'Reference_Node' })<-[:ROOT]->(root), p=(root)<-[PART_OF*0..]-()<-[:MEMBER_OF]-(user)
RETURN user.name, min(length(p))
ORDER BY min(length(p)), user.name
and results in the following output:

75

user.name
min(length(p))
"Ali"
"Burcu"
"Can"
"Engin"
"Demet"
"Fuat"
"Gul"
"Hakan"
"Irmak"
"Jale"
10 rows
in Java:
.breadthFirst()
.evaluator(
Evaluators.includeWhereLastRelationshipTypeIs( RoleRels.MEMBER_OF ) );
traverser = traversalDescription.traverse( referenceNode );
Found:
Found:
Found:
Found:
Found:
Found:
Found:
Found:
Found:
Found:
Ali at depth: 1
Engin at depth: 1
Burcu at depth: 1
Can at depth: 1
Demet at depth: 2
Gul at depth: 2
Fuat at depth: 2
Hakan at depth: 2
Irmak at depth: 2
Jale at depth: 3
As seen above, querying even more complex scenarios can be done using comparatively short
constructs in Cypher or Java.
76
Chapter6.Languages
Please see http://www.neo4j.org/drivers for the current set of drivers!
Theres an included Java example which shows a low-level approach to using the Neo4j REST API
from Java.
77
Languages
6.1.How to use the REST API from Java

6.1.1.Creating a graph through the REST API from Java
The REST API uses HTTP and JSON, so that it can be used from many languages and platforms.
Still, when geting started its useful to see some patterns that can be re-used. In this brief overview,
well show you how to create and manipulate a simple graph through the REST API and also how to
query it.
For these examples, weve chosen the Jersey <http://jersey.java.net/> client components, which are
easily downloaded <http://jersey.java.net/nonav/documentation/latest/user-guide.html#chapter_deps>
via Maven.
6.1.2.Start the server

Before we can perform any actions on the server, we need to start it as per Section21.2, Server
Installation.
WebResource resource = Client.create()
.resource( SERVER_ROOT_URI );
ClientResponse response = resource.get( ClientResponse.class );
System.out.println( String.format( "GET on [%s], status code [%d]",
SERVER_ROOT_URI, response.getStatus() ) );
response.close();
If the status of the response is 200 OK, then we know the server is running fine and we can continue. If
the code fails to conenct to the server, then please have a look at PartV, Operations.
Note
If you get any other response than 200 OK (particularly 4xx or 5xx responses) then please
check your configuration and look in the log files in the data/log directory.
6.1.3.Creating a node
The REST API uses POST to create nodes. Encapsulating that in Java is straightforward using the
Jersey client:
final String nodeEntryPointUri = SERVER_ROOT_URI + "node";
// http://localhost:7474/db/data/node
.resource( nodeEntryPointUri );
// POST {} to the node entry point URI
ClientResponse response = resource.accept( MediaType.APPLICATION_JSON )
.type( MediaType.APPLICATION_JSON )
.entity( "{}" )
.post( ClientResponse.class );
final URI location = response.getLocation();
System.out.println( String.format(
"POST to [%s], status code [%d], location header [%s]",
nodeEntryPointUri, response.getStatus(), location.toString() ) );
response.close();
return location;
If the call completes successfully, under the covers it will have sent a HTTP request containing
a JSON payload to the server. The server will then have created a new node in the database and
responded with a 201 Created response and a Location header with the URI of the newly created node.
78
Languages
In our example, we call this functionality twice to create two nodes in our database.
6.1.4.Adding properties
Once we have nodes in our datatabase, we can use them to store useful data. In this case, were going
to store information about music in our database. Lets start by looking at the code that we use to
create nodes and add properties. Here weve added nodes to represent "Joe Strummer" and a band
called "The Clash".
URI firstNode = createNode();
addProperty( firstNode, "name", "Joe Strummer" );
URI secondNode = createNode();
addProperty( secondNode, "band", "The Clash" );
Inside the addProperty method we determine the resource that represents properties for the node and
decide on a name for that property. We then proceed to PUT the value of that property to the server.
String propertyUri = nodeUri.toString() + "/properties/" + propertyName;
// http://localhost:7474/db/data/node/{node_id}/properties/{property_name}
.resource( propertyUri );
.entity( "\"" + propertyValue + "\"" )
.put( ClientResponse.class );
System.out.println( String.format( "PUT to [%s], status code [%d]",
propertyUri, response.getStatus() ) );
response.close();
If everything goes well, well get a 204 No Content back indicating that the server processed the request
but didnt echo back the property value.
6.1.5.Adding relationships
Now that we have nodes to represent Joe Strummer and The Clash, we can relate them. The REST
API supports this through a POST of a relationship representation to the start node of the relationship.
Correspondingly in Java we POST some JSON to the URI of our node that represents Joe Strummer, to
establish a relationship between that node and the node representing The Clash.
URI relationshipUri = addRelationship( firstNode, secondNode, "singer",
"{ \"from\" : \"1976\", \"until\" : \"1986\" }" );
Inside the addRelationship method, we determine the URI of the Joe Strummer nodes relationships,
and then POST a JSON description of our intended relationship. This description contains the
destination node, a label for the relationship type, and any attributes for the relation as a JSON
collection.
private static URI addRelationship( URI startNode, URI endNode,
String relationshipType, String jsonAttributes )
throws URISyntaxException
{
URI fromUri = new URI( startNode.toString() + "/relationships" );
String relationshipJson = generateJsonRelationship( endNode,
relationshipType, jsonAttributes );
.resource( fromUri );
// POST JSON to the relationships URI
79
Languages
.entity( relationshipJson )
final URI location = response.getLocation();
"POST to [%s], status code [%d], location header [%s]",
fromUri, response.getStatus(), location.toString() ) );
response.close();
return location;
}
If all goes well, we receive a 201 Created status code and a Location header which contains a URI of
the newly created relation.
6.1.6.Add properties to a relationship

Like nodes, relationships can have properties. Since were big fans of both Joe Strummer and the
Clash, well add a rating to the relationship so that others can see hes a 5-star singer with the band.
addMetadataToProperty( relationshipUri, "stars", "5" );
Inside the addMetadataToProperty method, we determine the URI of the properties of the relationship
and PUT our new values (since its PUT it will always overwrite existing values, so be careful).
private static void addMetadataToProperty( URI relationshipUri,
String name, String value ) throws URISyntaxException
{
URI propertyUri = new URI( relationshipUri.toString() + "/properties" );
String entity = toJsonNameValuePairCollection( name, value );
.resource( propertyUri );
.entity( entity )
.put( ClientResponse.class );
"PUT [%s] to [%s], status code [%d]", entity, propertyUri,
response.getStatus() ) );
response.close();
}
Assuming all goes well, well get a 204 OK response back from the server (which we can check by
calling ClientResponse.getStatus()) and weve now established a very small graph that we can query.
6.1.7.Querying graphs
As with the embedded version of the database, the Neo4j server uses graph traversals to look for data
in graphs. Currently the Neo4j server expects a JSON payload describing the traversal to be POST-ed at
the starting node for the traversal (though this is likely to change in time to a GET-based approach).
To start this process, we use a simple class that can turn itself into the equivalent JSON, ready for
POST-ing to the server, and in this case weve hardcoded the traverser to look for all nodes with
outgoing relationships with the type "singer".
// TraversalDefinition turns into JSON to send to the Server
TraversalDefinition t = new TraversalDefinition();
t.setOrder( TraversalDefinition.DEPTH_FIRST );
t.setUniqueness( TraversalDefinition.NODE );
80
Languages
t.setMaxDepth( 10 );
t.setReturnFilter( TraversalDefinition.ALL );
t.setRelationships( new Relation( "singer", Relation.OUT ) );
Once we have defined the parameters of our traversal, we just need to transfer it. We do this by
determining the URI of the traversers for the start node, and then POST-ing the JSON representation of
the traverser to it.
URI traverserUri = new URI( startNode.toString() + "/traverse/node" );
.resource( traverserUri );
String jsonTraverserPayload = t.toJson();
.entity( jsonTraverserPayload )
"POST [%s] to [%s], status code [%d], returned data: "
+ System.getProperty( "line.separator" ) + "%s",
jsonTraverserPayload, traverserUri, response.getStatus(),
response.getEntity( String.class ) ) );
response.close();
Once that request has completed, we get back our dataset of singers and the bands they belong to:
[ {
"outgoing_relationships" : "http://localhost:7474/db/data/node/82/relationships/out",
"data" : {
"band" : "The Clash",
"name" : "Joe Strummer"
},
"traverse" : "http://localhost:7474/db/data/node/82/traverse/{returnType}",
"all_typed_relationships" : "http://localhost:7474/db/data/node/82/relationships/all/{-list|&|types}",
"property" : "http://localhost:7474/db/data/node/82/properties/{key}",
"all_relationships" : "http://localhost:7474/db/data/node/82/relationships/all",
"self" : "http://localhost:7474/db/data/node/82",
"properties" : "http://localhost:7474/db/data/node/82/properties",
"outgoing_typed_relationships" : "http://localhost:7474/db/data/node/82/relationships/out/{-list|&|types}",
"incoming_relationships" : "http://localhost:7474/db/data/node/82/relationships/in",
"incoming_typed_relationships" : "http://localhost:7474/db/data/node/82/relationships/in/{-list|&|types}",
"create_relationship" : "http://localhost:7474/db/data/node/82/relationships"
}, {
"data" : {
},
"create_relationship" : "http://localhost:7474/db/data/node/83/relationships"
} ]
6.1.8.Phew, is that it?

Thats a flavor of what we can do with the REST API. Naturally any of the HTTP idioms we provide
on the server can be easily wrapped, including removing nodes and relationships through DELETE.
Still if youve gotten this far, then switching .post() for .delete() in the Jersey client code should be
straightforward.
81
Languages
6.1.9.Whats next?
The HTTP API provides a good basis for implementers of client libraries, its also great for HTTP
and REST folks. In the future though we expect that idiomatic language bindings will appear to take
advantage of the REST API while providing comfortable language-level constructs for developers to
use, much as there are similar bindings for the embedded database.
6.1.10.Appendix: the code

CreateSimpleGraph.java <https://github.com/neo4j/neo4j/blob/2.0.0/community/server-examples/
src/main/java/org/neo4j/examples/server/CreateSimpleGraph.java>
Relation.java <https://github.com/neo4j/neo4j/blob/2.0.0/community/server-examples/src/main/
java/org/neo4j/examples/server/Relation.java>
TraversalDefinition.java <https://github.com/neo4j/neo4j/blob/2.0.0/community/server-examples/
src/main/java/org/neo4j/examples/server/TraversalDefinition.java>
82
PartIII.Cypher Query Language
The Cypher part is the authoritative source for details on the Cypher Query Language. For an
introduction, see Section7.1, What is Cypher?.
Chapter7.Introduction
To get an overview of Cypher, continue reading Section7.1, What is Cypher?. The rest of this
chapter deals with the context of Cypher statements, like for example transaction management and
how to use parameters. For the Cypher language reference itself see other chapters at PartIII, Cypher
Query Language.
84
Introduction
7.1.What is Cypher?
7.1.1.Introduction
Cypher is a declarative graph query language that allows for expressive and efficient querying and
updating of the graph store. Cypher is a relatively simple but still very powerful language. Very
complicated database queries can easily be expressed through Cypher. This allows you to focus on
your domain instead of getting lost in database access.
Cypher is designed to be a humane query language, suitable for both developers and (importantly, we
think) operations professionals. Our guiding goal is to make the simple things easy, and the complex
things possible. Its constructs are based on English prose and neat iconography which helps to make
queries more self-explanatory. We have tried to optimize the language for reading and not for writing.
Being a declarative language, Cypher focuses on the clarity of expressing what to retrieve from
a graph, not on how to retrieve it. This is in contrast to imperative languages like Java, scripting
languages like Gremlin <http://gremlin.tinkerpop.com>, and the JRuby Neo4j bindings <http://
neo4j.rubyforge.org/>. This approach makes query optimization an implementation detail instead of
burdening the user with it and requiring her to update all traversals just because the physical database
structure has changed (new indexes etc.).
Cypher is inspired by a number of different approaches and builds upon established practices for
expressive querying. Most of the keywords like WHERE and ORDER BY are inspired by SQL <http://
en.wikipedia.org/wiki/SQL>. Pattern matching borrows expression approaches from SPARQL
<http://en.wikipedia.org/wiki/SPARQL>. Some of the collection semantics have been borrowed from
languages such as Haskell and Python.
7.1.2.Structure
Cypher borrows it structure from SQL queries are built up using various clauses.
Clauses are chained together, and the they feed intermediate result sets between each other. For
example, the matching identifiers from one MATCH clause will be the context that the next clause exists
in.
The query language is comprised of several distinct clauses. You can read more details about them
later in the manual.
Here are a few clauses used to read from the graph:
MATCH: The graph pattern to match. This is the most common way to get data from the graph.
WHERE: Not a clause in its own right, but rather part of MATCH, OPTIONAL MATCH and WITH. Adds
constraints to a pattern, or filters the intermediate result passing through WITH.
RETURN: What to return.
Lets see MATCH and RETURN in action.
Imagine an example graph like the following one:
85
Introduction
Figure7.1.Example Graph
Node[ 3] nam e = 'John'
friend
friend
Node[ 0] nam e = 'Sara'
friend
friend
Node[ 1] nam e = 'Maria'
Node[ 2] nam e = 'St eve'
For example, here is a query which finds a user called John and Johns friends (though not his direct
friends) before returning both John and any friends-of-friends that are found.
MATCH (john {name: 'John'})-[:friend]->()-[:friend]->(fof)
RETURN john, fof
Resulting in:
john
fof
Node[3]{name:"John"}
Node[1]{name:"Maria"}
Node[2]{name:"Steve"}
2 rows
Next up we will add filtering to set more parts in motion:
We take a list of user names and find all nodes with names from this list, match their friends and
return only those followed users who have a name property starting with S.
MATCH (user)-[:friend]->(follower)
WHERE user.name IN ['Joe', 'John', 'Sara', 'Maria', 'Steve'] AND follower.name =~ 'S.*'
RETURN user, follower.name
Resulting in:
user
follower.name
"Sara"
Node[4]{name:"Joe"}
"Steve"
2 rows
And here are examples of clauses that are used to update the graph:
CREATE (and DELETE): Create (and delete) nodes and relationships.
SET (and REMOVE): Set values to properties and add labels on nodes using SET and use REMOVE to
remove them.
86
Introduction
MERGE: Match existing or create new nodes and patterns. This is especially useful together with
uniqueness constraints.
For more Cypher examples, see Chapter5, Data Modeling Examples as well as the rest of the Cypher
part with details on the language. To use Cypher from Java, see Section32.12, Execute Cypher
Queries from Java.
87
Introduction
7.2.Updating the graph

Cypher can be used for both querying and updating your graph.
7.2.1.The Structure of Updating Queries

Quick info
A Cypher query part cant both match and update the graph at the same time.
Every part can either read and match on the graph, or make updates on it.
If you read from the graph, and then update the graph, your query implicitly has two parts the
reading is the first part, and the writing is the second. If your query is read-only, Cypher will be lazy,
and not actually match the pattern until you ask for the results. In an updating query, the semantics are
that all the reading will be done before any writing actually happens. First reading, and then writing, is
the only pattern where the query parts are implicit any other order and you have to be explicit about
your query parts. The parts are separated using the WITH statement. WITH is like the event horizon its
a barrier between a plan and the finished execution of that plan.
When you want to filter using aggregated data, you have to chain together two reading query
parts the first one does the aggregating, and the second query filters on the results coming from the
first one.
START n=node(...)
MATCH n-[:friend]-friend
WITH n, count(friend) as friendsCount
WHERE friendsCount > 3
RETURN n, friendsCount
Using WITH, you specify how you want the aggregation to happen, and that the aggregation has to be
finished before Cypher can start filtering.
You can chain together as many query parts as you have JVM heap for.
7.2.2.Returning data
Any query can return data. If your query only reads, it has to return data it serves no purpose if it
doesnt, and it is not a valid Cypher query. Queries that update the graph dont have to return anything,
but they can.
After all the parts of the query comes one final RETURN clause. RETURN is not part of any query part it
is a period symbol at the end of a query. The RETURN clause has three sub-clauses that come with it
SKIP/LIMIT and ORDER BY.
If you return graph elements from a query that has just deleted them beware, you are holding a
pointer that is no longer valid. Operations on that node might fail mysteriously and unpredictably.
88
Introduction
7.3.Transactions
Any query that updates the graph will run in a transaction. An updating query will always either fully
succeed, or not succeed at all.
Cypher will either create a new transaction or run inside an existing one:
If no transaction exists in the running context Cypher will create one and commit it once the query
finishes.
In case there already exists a transaction in the running context, the query will run inside it, and
nothing will be persisted to disk until that transaction is successfully committed.
This can be used to have multiple queries be committed as a single transaction:
1. Open a transaction,
2. run multiple updating Cypher queries,
3. and commit all of them in one go.
Note that a query will hold the changes in memory until the whole query has finished executing. A
large query will consequently need a JVM with lots of heap space.
For using transactions over the REST API, see Section19.1, Transactional HTTP endpoint.
When using Neo4j embedded, remember all iterators returned from an execution result should be
exhausted fully to ensure that resources bound to them will be properly closed. Resources include
transactions started by the query, so failing to do so may, for example, lead to deadlocks or other
weird behavior.
89
Introduction
7.4.Uniqueness
While pattern matching, Cypher makes sure to not include matches where the same graph relationship
is found multiple times in a single pattern. In most use cases, this is a sensible thing to do.
Example: looking for a users friends of friends should not return said user.
Lets create a few nodes and relationships:
CREATE (adam:User { name: 'Adam' }),(pernilla:User { name: 'Pernilla' }),(david:User { name: 'David'
}),
(adam)-[:FRIEND]->(pernilla),(pernilla)-[:FRIEND]->(david)
Which gives us the following graph:

User
nam e = 'Adam '
FRIEND
User
nam e = 'Pernilla'
FRIEND
User
nam e = 'David'
Now lets look for friends of friends of Adam:

MATCH (user:User { name: 'Adam' })-[r1:FRIEND]-()-[r2:FRIEND]-(friend_of_a_friend)
RETURN friend_of_a_friend
friend_of_a_friend
Node[2]{name:"David"}
1 row
In this query, Cypher makes sure to not return matches where the pattern relationships r1 and r2 point
to the same graph relationship.
This is however not always desired. If the query should return the user, it is possible to spread the
matching over multiple MATCH clauses, like so:
MATCH (user:User { name: 'Adam' })-[r1:FRIEND]-(friend)
WITH friend
MATCH (friend)-[r2:FRIEND]-(friend_of_a_friend)
RETURN friend_of_a_friend
friend_of_a_friend
Node[2]{name:"David"}
2 rows
90
Introduction
friend_of_a_friend
Node[0]{name:"Adam"}
2 rows
91
Introduction
7.5.Parameters
Cypher supports querying with parameters. This means developers dont have to resort to string
building to create a query. In addition to that, it also makes caching of execution plans much easier for
Cypher.
Parameters can be used for literals and expressions in the WHERE clause, for the index value in the
START clause, index queries, and finally for node/relationship ids. Parameters can not be used as for
property names, relationship types and labels, since these patterns are part of the query structure that is
compiled into a query plan.
Accepted names for parameters are letters and numbers, and any combination of these.
For details on parameters when using the Neo4j embedded Java API, see Section32.13, Query
Parameters. For details on using parameters via the Neo4j REST API, see Section19.5, Cypher
queries via REST.
Below follows a comprehensive set of examples of parameter usage. The parameters are given as
JSON here. Exactly how to submit them depends on the driver in use.
7.5.1.String literal
Parameters.
{
"name" : "Johan"
}
Query.
MATCH (n)
WHERE n.name = { name }
RETURN n
7.5.2.Regular expression
Parameters.
{
"regex" : ".*h.*"
}
Query.
MATCH (n)
WHERE n.name =~ { regex }
RETURN n.name
7.5.3.Create node with properties

Parameters.
{
"props" : {
"position" : "Developer",
"name" : "Andres"
}
}
Query.
CREATE ({ props })
92
Introduction
7.5.4.Create multiple nodes with properties

Parameters.
{
"props" : [ {
"awesome" : true,
"name" : "Andres"
}, {
"name" : "Michael",
"children" : 3
} ]
}
Query.
CREATE (n:Person { props })
RETURN n
7.5.5.Setting all properties on node

Note that this will replace all the current properties.
Parameters.
{
"props" : {
"name" : "Andres"
}
}
Query.
MATCH (n)
WHERE n.name='Michaela'
SET n = { props }
7.5.6.SKIP and LIMIT

Parameters.
{
"s" : 1,
"l" : 1
}
Query.
MATCH (n)
RETURN n.name
SKIP { s }
LIMIT { l }
7.5.7.Node id
Parameters.
{
"id" : 0
}
Query.
START n=node({ id })
93
Introduction
RETURN n.name
7.5.8.Multiple node ids

Parameters.
{
"id" : [ 0, 1, 2 ]
}
Query.
START n=node({ id })
RETURN n.name
7.5.9.Index value (legacy indexes)

Parameters.
{
"value" : "Michaela"
}
Query.
START n=node:people(name = { value })
RETURN n
7.5.10.Index query (legacy indexes)

Parameters.
{
"query" : "name:Andreas"
}
Query.
START n=node:people({ query })
RETURN n
94
Introduction
7.6.Compatibility
Cypher is still changing rather rapidly. Parts of the changes are internal we add new pattern
matchers, aggregators and other optimizations, which hopefully makes your queries run faster.
Other changes are directly visible to our users the syntax is still changing. New concepts are being
added and old ones changed to fit into new possibilities. To guard you from having to keep up with
our syntax changes, Cypher allows you to use an older parser, but still gain the speed from new
optimizations.
There are two ways you can select which parser to use. You can configure your database with the
configuration parameter cypher_parser_version, and enter which parser youd like to use (1.9, 2.0 are
supported now). Any Cypher query that doesnt explicitly say anything else, will get the parser you
have configured.
The other way is on a query by query basis. By simply putting "CYPHER 1.9" at the beginning, that
particular query will be parsed with the 1.9 version of the parser. Example:
CYPHER 1.9 START n=node(0)
WHERE n.foo = "bar"
RETURN n
95
Introduction
7.7.Query Performance
Cypher works very hard to execute queries as fast as possible.
However, when optimizing for maximum query execution performance, it may be helpful to rephrase
queries using knowledge about the domain and the application.
The overall goal of manual query performance optimization is to ensure that only necessary data is
retrieved from the graph. At least data should get filtered out as early as possible in order to reduce
the amount of work that has to be done at later stages of query execution. This also goes for what gets
returned: avoid returning whole nodes and relationships instead, pick the data you need and return
only that. You should also make sure to set an upper limit on variable length patterns, so they dont
cover larger portions of the dataset than needed.
Each Cypher query gets optimized and transformed into an execution plan by the Cypher execution
engine. To minimize the resources used for this, make sure to use parameters instead of literals
when possible. This allows Cypher to re-use your queries instead of having to parse and build new
execution plans.
Note
When Cypher is building execution plans, it looks at the schema to see if it can find useful
indexes. These index decisions are only valid until the schema changes, so adding or
removing indexes leads to the execution plan cache being flushed.
96
Chapter8.Syntax
The nitty-gritty details of Cypher syntax.
97
Syntax
8.1.Operators
8.1.1.Mathematical operators
The mathematical operators are +, -, *, / and %, ^.
8.1.2.Comparison operators
The comparison operators are =, <>, <, >, <=, >=.
8.1.3.Boolean operators
The boolean operators are AND, OR, XOR, NOT.
8.1.4.String operators
Strings can be concatenated using the + operator.
8.1.5.Collection operators
Collections can be concatenated using the + operator. To check if an element exists in a collection, you
can use the IN operator.
8.1.6.Property operators
Note
Since version 2.0, the previously existing property operators ? and ! have been removed.
This syntax is no longer supported. Missing properties are now returned as null. Please use
(NOT(has(<ident>.prop)) OR <ident>.prop=<value>) if you really need the old behavior of the
? operator. Also, the use of ? for optional relationships has been removed in favor of the
newly introduced OPTIONAL MATCH clause.
98
Syntax
8.2.Expressions
8.2.1.Expressions in general
An expression in Cypher can be:
A numeric literal (integer or double): 13, 40000, 3.14.

A string literal: "Hello", 'World'.
A boolean literal: true, false, TRUE, FALSE.
An identifier: n, x, rel, myFancyIdentifier, À name with weird stuff in it[]!`.
A property: n.prop, x.prop, rel.thisProperty, myFancyIdentifier.`(weird property name)`.
A parameter: {param}, {0}
A collection of expressions: ["a", "b"], [1,2,3], ["a", 2, n.property, {param}], [ ].
A function call: length(p), nodes(p).
An aggregate function: avg(x.prop), count(*).
A path-pattern: (a)-->()<--(b).
A predicate expression is an expression that returns true or false: a.prop = "Hello", length(p) > 10,
has(a.name)
a CASE expression
8.2.2.Note on string literals

String literals can contain these escape sequences.
Escape Character
sequence
\t
Tab
\b
Backspace
\n
Newline
\r
Carriage return
\f
Form feed
\'
Single quote
\"
Double quote
\\
Backslash
8.2.3.Case Expressions
Cypher supports CASE expressions, which is a generic conditional expression, similar to if/else
statements in other languages. Two variants of CASE exist the simple form and the generic form.
8.2.4.Simple CASE
The expression is calculated, and compared in order with the WHEN clauses until a match is found. If no
match is found the expression in the ELSE clause is used, or null, if no ELSE case exists.
Syntax:
CASE test
WHEN value THEN result
99
Syntax
[WHEN ...]
[ELSE default]
END
Arguments:
expr: A valid expression.

value: An expression whose result will be compared to the test expression.
result: This is the result expression used if the value expression matches the test expression.
default: The expression to use if no match is found.
Query.
MATCH n
RETURN CASE n.eyes
WHEN 'blue'
THEN 1
WHEN 'brown'
THEN 2
ELSE 3 END AS result
Result
result
2
1
2
1
3
5 rows
8.2.5.Generic CASE
The predicates are evaluated in order until a true value is found, and the result value is used. If no
match is found the expression in the ELSE clause is used, or null, if no ELSE case exists.
Syntax:
CASE
WHEN predicate THEN result
[WHEN ...]
[ELSE default]
END
Arguments:
predicate: A predicate that is tested to find a valid alternative.
result: This is the result expression used if the predicate matches.
default: The expression to use if no match is found.
Query.
MATCH n
RETURN CASE
WHEN n.eyes = 'blue'
THEN 1
100
Syntax
WHEN n.age < 40
THEN 2
ELSE 3 END AS result
Result
result
3
1
2
1
3
5 rows
101
Syntax
8.3.Identifiers
When you reference parts of the pattern, you do so by naming them. The names you give the different
parts are called identifiers.
In this example:
MATCH (n)-->(b) RETURN b
The identifiers are n and b.

Identifier names are case sensitive, and can contain underscores and alphanumeric characters (a-z,
0-9), but must start with a letter. If other characters are needed, you can quote the identifier using
backquote (`) signs.
The same rules apply to property names.
102
Syntax
8.4.Comments
To add comments to your queries, use double slash. Examples:
MATCH (n) RETURN n //This is an end of line comment
MATCH (n)
//This is a whole line comment
RETURN n
MATCH (n) WHERE n.property = "//This is NOT a comment" RETURN n
103
Syntax
8.5.Patterns
Patterns and pattern-matching are at the very heart of Cypher, so being effective with Cypher requires
a good understanding of patterns.
Using patterns, you describe the shape of the data youre looking for. For example, in the MATCH clause
you describe the shape with a pattern, and Cypher will figure out how to get that data for you.
The pattern describes the data using a form that is very similar to how one typically draws the shape
of property graph data on a whiteboard: usually as circles (representing nodes) and arrows between
them to represent relationships.
Patterns appear in multiple places in Cypher: in MATCH, CREATE and MERGE clauses, and in pattern
expressions. Each of these is described in more details in:
Section10.1, Match
Section10.2, Optional Match
Section11.1, Create
Section11.3, Merge
Section10.3.3, Using patterns in WHERE
8.5.1.Patterns for nodes

The very simplest shape that can be described in a pattern is a node. A node is described using a
pair of parentheses, and is typically given a name. For example:
(a)
This simple pattern describes a single node, and names that node using the identifier a.
Note that the parentheses may be omitted, but only when there are no labels or properties specified for
the node pattern.
8.5.2.Patterns for related nodes

More interesting is patterns that describe multiple nodes and relationships between them. Cypher
patterns describe relationships by employing an arrow between two nodes. For example:
(a)-->(b)
This pattern describes a very simple data shape: two nodes, and a single relationship from one to the
other. In this example, the two nodes are both named as a and b respectively, and the relationship is
directed: it goes from a to b.
This way of describing nodes and relationships can be extended to cover an arbitrary number of nodes
and the relationships between them, for example:
(a)-->(b)<--(c)
Such a series of connected nodes and relationships is called a "path".

Note that the naming of the nodes in these patterns is only necessary should one need to refer to the
same node again, either later in the pattern or elsewhere in the Cypher query. If this is not necessary
then the name may be omitted, like so:
(a)-->()<--(c)
104
Syntax
8.5.3.Labels
In addition to simply describing the shape of a node in the pattern, one can also describe attributes.
The most simple attribute that can be described in the pattern is a label that the node must have. For
example:
(a:User)-->(b)
One can also describe a node that has multiple labels:

(a:User:Admin)-->(b)
8.5.4.Specifying properties
Nodes and relationships are the fundamental structures in a graph. Neo4j uses properties on both of
these to allow for far richer models.
Properties can be expressed in patterns using a map-construct: curly brackets surrounding a number of
key-expression pairs, separated by commas. E.g. a node with two properties on it would look like: (a
{ name: "Andres", sport: "Brazilian Ju-Jitsu" }).
A relationship with expectations on it would could look like: (a)-[{blocked: false}]->(b).

When properties appear in patterns, they add an additional constraint to the shape of the data. In the
case of a CREATE clause, the properties will be set in the newly created nodes and relationships. In the
case of a MERGE clause, the properties will be used as additional constraints on the shape any existing
data must have (the specified properties must exactly match any existing data in the graph). If no
matching data is found, then MERGE behaves like CREATE and the properties will be set in the newly
created nodes and relationships.
Note that patterns supplied to CREATE may use a single parameter to specify properties, e.g: CREATE
(node {paramName}). This is not possible with patterns used in other clauses, as Cypher needs to know
the property names at the time the query is compiled, so that matching can be done effectively.
8.5.5.Describing relationships
The simplest way to describe a relationship is by using the arrow between two nodes, as in the
previous examples. Using this technique, you can describe that the relationship should exist and the
directionality of it. If you dont care about the direction of the relationship, the arrow head can be
omitted, like so:
(a)--(b)
As with nodes, relationships may also be given names. In this case, a pair of square brackets is used to
break up the arrow and the identifier is placed between. For example:
(a)-[r]->(b)
Much like labels on nodes, relationships can have types. To describe a relationship with a specific
type, you can specify this like so:
(a)-[r:REL_TYPE]->(b)
Unlike labels, relationships can only have one type. But if wed like to describe some data such
that the relationship could have any one of a set of types, then they can all be listed in the pattern,
separating them with the pipe symbol | like this:
(a)-[r:TYPE1|TYPE2]->(b)
Note that this form of pattern can only be used to describe existing data (ie. when using a pattern
with MATCH or as an expression). It will not work with CREATE or MERGE, since its not possible to create a
relationship with multiple types.
105
Syntax
As with nodes, the name of the relationship can always be omitted, in this case like so:
(a)-[:REL_TYPE]->(b)
Variable length
Rather than describing a long path using a sequence of many node and relationship descriptions in a
pattern, many relationships (and the intermediate nodes) can be described by specifying a length in the
relationship description of a pattern. For example:
(a)-[*2]->(b)
This describes a graph of three nodes and two relationship, all in one path (a path of length 2). This is
equivalent to:
(a)-->()-->(b)
A range of lengths can also be specified: such relationship patterns are called variable length
relationships. For example:
(a)-[*3..5]->(b)
This is a minimum length of 3, and a maximum of 5. It describes a graph of either 4 nodes and 3
relationships, 5 nodes and 4 relationships or 6 nodes and 5 relationships, all connected together in a
single path.
Either bound can be omitted. For example, to describe paths of length 3 or more, use:
(a)-[*3..]->(b)
And to describe paths of length 5 or less, use:

(a)-[*..5]->(b)
Both bounds can be omitted, allowing paths of any length to be described:

(a)-[*]->(b)
As a simple example, lets take the query below:

Query.
MATCH (me)-[:KNOWS*1..2]-(remote_friend)
WHERE me.name = "Filipa"
RETURN remote_friend.name
Result
remote_friend.name
"Dilshad"
"Anders"
2 rows
This query finds data in the graph which a shape that fits the pattern: specifically a node (with the
name property Filipa) and then the KNOWS related nodes, one or two steps out. This is a typical example
of finding first and second degree friends.
Note that variable length relationships can not be used with CREATE and MERGE.
8.5.6.Assigning to path identifiers

As described above, a series of connected nodes and relationships is called a "path". Cypher allows
paths to be named using an identifer, like so:
106
Syntax
p = (a)-[*3..5]->(b)
You can do this in MATCH, CREATE and MERGE, but not when using patterns as expressions.
107
Syntax
8.6.Collections
Cypher has good support for collections.
8.6.1.Collections in general
A literal collection is created by using brackets and separating the elements in the collection with
commas.
Query.
RETURN [0,1,2,3,4,5,6,7,8,9] AS collection
Result
collection
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
1 row
In our examples, well use the range function. It gives you a collection containing all numbers
between given start and end numbers.
To access individual elements in the collection, we use the square brackets again.
Query.
RETURN range(0,10)[3]
Result
range(0,10)[3]
3
1 row
You can also use negative numbers, to start from the end of the collection instead.
Query.
RETURN range(0,10)[-3]
Result
range(0,10)[-3]
8
1 row
Finally, you can use ranges inside the brackets to return ranges of the collection.
Query.
RETURN range(0,10)[0..3]
Result
range(0,10)[0..3]
[0, 1, 2]
1 row
Query.
108
Syntax
RETURN range(0,10)[0..-5]
Result
range(0,10)[0..-5]
[0, 1, 2, 3, 4, 5]
1 row
Query.
RETURN range(0,10)[-5..]
Result
range(0,10)[-5..]
[6, 7, 8, 9, 10]
1 row
Query.
RETURN range(0,10)[..4]
Result
range(0,10)[..4]
[0, 1, 2, 3]
1 row
Note: Out-of-bound slices are simply truncated, but out-of-bound single elements return null.
Query.
RETURN range(0,10)[15]
Result
range(0,10)[15]
<null>
1 row
Query.
RETURN range(0,10)[5..15]
Result
range(0,10)[5..15]
[5, 6, 7, 8, 9, 10]
1 row
8.6.2.List comprehension
List comprehension is a syntactic construct available in Cypher for creating a collection based on
existing collections. It follows the form of the mathematical set-builder notation (set comprehension)
instead of the use of map and filter functions.
Query.
109
Syntax
RETURN [x IN range(0,10) WHERE x % 2 = 0 | x^3] AS result
Result
result
[0. 0, 8. 0, 64. 0, 216. 0, 512. 0, 1000. 0]
1 row
Either the WHERE part, or the expression, can be omitted, if you only want to filter or map respectively.
Query.
RETURN [x IN range(0,10) WHERE x % 2 = 0] AS result
Result
result
[0, 2, 4, 6, 8, 10]
1 row
Query.
RETURN [x IN range(0,10)| x^3] AS result
Result
result
[0. 0, 1. 0, 8. 0, 27. 0, 64. 0, 125. 0, 216. 0, 343. 0, 512. 0, 729. 0, 1000. 0]
1 row
8.6.3.Literal maps
From Cypher, you can also construct maps. In Java, they will be java.util.Map<String,Object>,
through REST you will get JSON objects.
Query.
RETURN { key : "Value", collectionKey: [{ inner: "Map1" }, { inner: "Map2" }]}
Result
{ key : "Value", collectionKey: [ { inner: "Map1" }, { inner: "Map2" } ] }
{key -> "Value", collectionKey -> [{inner -> "Map1"}, {inner -> "Map2"}]}
1 row
110
Syntax
8.7.Working with NULL

In Cypher, NULL is used to represent missing or undefined values. Conceptually, NULL means a missing
unknown value and it is treated somewhat differently from other values. E.g. getting a property from
a node that does not have said property produces NULL. Most expressions that take NULL as input will
produce NULL. This includes boolean expressions that are used as predicates in the WHERE clause. In this
case, anything that is not TRUE is interpreted as being false.
is not equal to NULL. Not knowing two values does not imply that they are the same value. So the
expression NULL = NULL yields NULL and not TRUE.
NULL
The logical operators (i.e. AND, OR, XOR, IN) treat NULL as the unknown value of three-valued logic.
Here is the truth table for AND, OR and XOR.
a
a AND b
a OR b
a XOR b
FALSE
FALSE
FALSE
FALSE
FALSE
FALSE
NULL
FALSE
NULL
NULL
FALSE
TRUE
FALSE
TRUE
TRUE
TRUE
FALSE
FALSE
TRUE
TRUE
TRUE
NULL
NULL
TRUE
NULL
TRUE
TRUE
TRUE
TRUE
FALSE
NULL
FALSE
FALSE
NULL
NULL
NULL
NULL
NULL
NULL
NULL
NULL
TRUE
NULL
TRUE
NULL
The IN operator follows similar logic. If Cypher knows that something exists in a collection, the result
will be TRUE. Any collection that contains a NULL and doesnt have a matching element will return NULL.
Otherwise, the result will be false. Here is a table with examples:
Expression
Result
2 IN [1, 2, 3]
TRUE
2 IN [1, NULL, 3]
NULL
2 IN [1, 2, NULL]
TRUE
2 IN [1]
FALSE
2 IN []
FALSE
IN [1,2,3]
NULL
IN [1,NULL,3]
NULL
NULL
NULL
NULL IN
[]
FALSE
Using ALL, ANY, NONE, and SINGLE follows a similar rule. If the result can be calculated definitely, TRUE or
FALSE is returned. Otherwise NULL is produced.
8.7.1.Expressions that return NULL

Getting a missing element from a collection: [][0], head([])
111
Syntax
Trying to access a property that does not exist on a node or relationship: n.missingProperty
Comparisons when either side is NULL: 1 < NULL
Arithmetic expressions containing NULL: 1 + NULL
Function calls where any arguments are NULL: sin(NULL)
112
Chapter9.General Clauses
113
General Clauses
9.1.Return
In the RETURN part of your query, you define which parts of the pattern you are interested in. It can be
nodes, relationships, or properties on these.
Figure9.1.Graph
nam e = 'A'
happy = 'Yes! '
age = 55
KNOWS BLOCKS
nam e = 'B'
9.1.1.Return nodes
To return a node, list it in the RETURN statement.
Query.
MATCH (n { name: "B" })
RETURN n
The example will return the node.

Result
n
Node[1]{name:"B"}
1 row
9.1.2.Return relationships
To return a relationship, just include it in the RETURN list.
Query.
MATCH (n { name: "A" })-[r:KNOWS]->(c)
RETURN r
The relationship is returned by the example.

Result
r
:KNOWS[0]{}
1 row
9.1.3.Return property
To return a property, use the dot separator, like this:
Query.
MATCH (n { name: "A" })
RETURN n.name
114
General Clauses
The value of the property name gets returned.
Result
n.name
"A"
1 row
9.1.4.Return all elements

When you want to return all nodes, relationships and paths found in a query, you can use the *
symbol.
Query.
MATCH p=(a { name: "A" })-[r]->(b)
RETURN *
This returns the two nodes, the relationship and the path used in the query.
Result
b
Node[1]{name:"B"}
Node[0]{name:"A",
happy:"Yes!", age:55}
:KNOWS[0]{}
[Node[0]{name:"A",
happy:"Yes!",
age:55}, :KNOWS[0]{},
Node[1]{name:"B"}]
Node[1]{name:"B"}
Node[0]{name:"A",
happy:"Yes!", age:55}
:BLOCKS[1]{}
[Node[0]{name:"A",
happy:"Yes!",
age:55}, :BLOCKS[1]{},
Node[1]{name:"B"}]
2 rows
9.1.5.Identifier with uncommon characters

To introduce a placeholder that is made up of characters that are outside of the english alphabet, you
can use the ` to enclose the identifier, like this:
Query.
MATCH (`This isn't a common identifier`)
WHERE `This isn't a common identifier`.name='A'
RETURN `This isn't a common identifier`.happy
The node indexed with name "A" is returned

Result
`This isn't a common identifier`.happy
"Yes!"
1 row
9.1.6.Column alias
If the name of the column should be different from the expression used, you can rename it by using AS
<new name>.
115
General Clauses
Query.
MATCH (a { name: "A" })
RETURN a.age AS SomethingTotallyDifferent
Returns the age property of a node, but renames the column.

Result
SomethingTotallyDifferent
55
1 row
9.1.7.Optional properties
If a property might or might not be there, you can still select it as usual. It will be treated as NULL if it is
missing
Query.
MATCH (n)
RETURN n.age
This example returns the age when the node has that property, or null if the property is not there.
Result
n.age
55
<null>
2 rows
9.1.8.Other expressions
Any expression can be used as a return item literals, predicates, properties, functions, and
everything else.
Query.
MATCH (a { name: "A" })
RETURN a.age > 30, "I'm a literal",(a)-->()
Returns a predicate, a literal and function call with a pattern expression parameter.
Result
a.age > 30
"I'm a literal"
(a)-->()
true
"I'm a literal"
[[Node[0]{name:"A",
happy:"Yes!", age:55}, :KNOWS[0]
{}, Node[1]{name:"B"}], [Node[0]
{name:"A", happy:"Yes!",
age:55}, :BLOCKS[1]{}, Node[1]
{name:"B"}]]
1 row
9.1.9.Unique results
DISTINCT
retrieves only unique rows depending on the columns that have been selected to output.
116
General Clauses
Query.
MATCH (a { name: "A" })-->(b)
RETURN DISTINCT b
The node named B is returned by the query, but only once.

Result
b
Node[1]{name:"B"}
1 row
117
General Clauses
9.2.Order by
To sort the output, use the ORDER BY clause. Note that you can not sort on nodes or relationships, just
on properties on these.
Figure9.2.Graph
nam e = 'A'
age = 34
lengt h = 170
KNOWS
nam e = 'B'
age = 34
KNOWS
nam e = 'C'
age = 32
lengt h = 185
9.2.1.Order nodes by property

ORDER BY
is used to sort the output.
Query.
MATCH (n)
RETURN n
ORDER BY n.name
The nodes are returned, sorted by their name.

Result
n
Node[0]{name:"A", age:34, length:170}
Node[1]{name:"B", age:34}
Node[2]{name:"C", age:32, length:185}
3 rows
9.2.2.Order nodes by multiple properties

You can order by multiple properties by stating each identifier in the ORDER BY clause. Cypher will sort
the result by the first identifier listed, and for equals values, go to the next property in the ORDER BY
clause, and so on.
Query.
MATCH (n)
RETURN n
ORDER BY n.age, n.name
This returns the nodes, sorted first by their age, and then by their name.
118
General Clauses
Result
n
3 rows
9.2.3.Order nodes in descending order

By adding DESC[ENDING] after the identifier to sort on, the sort will be done in reverse order.
Query.
MATCH (n)
RETURN n
ORDER BY n.name DESC
The example returns the nodes, sorted by their name reversely.

Result
n
3 rows
9.2.4.Ordering NULL
When sorting the result set, NULL will always come at the end of the result set for ascending sorting,
and first when doing descending sort.
Query.
MATCH (n)
RETURN n.length, n
ORDER BY n.length
The nodes are returned sorted by the length property, with a node without that property last.
Result
n.length
170
185
<null>
3 rows
119
General Clauses
9.3.Limit
LIMIT
enables the return of only subsets of the total result.

Figure9.3.Graph
nam e = 'A'
KNOWS
nam e = 'D'
KNOWS
nam e = 'E'
KNOWS
KNOWS
nam e = 'B'
9.3.1.Return first part

To return a subset of the result, starting from the top, use this syntax:
Query.
MATCH (n)
RETURN n
LIMIT 3
The top three items are returned by the example query.

Result
n
Node[0]{name:"D"}
Node[1]{name:"E"}
Node[2]{name:"A"}
3 rows
120
nam e = 'C'
General Clauses
9.4.Skip
enables the return of only subsets of the total result. By using SKIP, the result set will get trimmed
from the top. Please note that no guarantees are made on the order of the result unless the query
specifies the ORDER BY clause.
SKIP
Figure9.4.Graph
nam e = 'A'
KNOWS
nam e = 'D'
KNOWS
nam e = 'E'
KNOWS
nam e = 'B'
KNOWS
nam e = 'C'
9.4.1.Skip first three

To return a subset of the result, starting from the fourth result, use the following syntax:
Query.
MATCH (n)
RETURN n
ORDER BY n.name
SKIP 3
The first three nodes are skipped, and only the last two are returned in the result.
Result
n
Node[0]{name:"D"}
Node[1]{name:"E"}
2 rows
9.4.2.Return middle two

To return a subset of the result, starting from somewhere in the middle, use this syntax:
Query.
MATCH (n)
RETURN n
ORDER BY n.name
SKIP 1
LIMIT 2
Two nodes from the middle are returned.

Result
n
Node[3]{name:"B"}
2 rows
121
General Clauses
n
Node[4]{name:"C"}
2 rows
122
General Clauses
9.5.With
With WITH, you can manipulate the result sequence before it is passed on to the following query parts.
The manipulations can be of the shape and/or number of entries in the result set.
One common usage of WITH is to limit the number of entries that are then passed on to other MATCH
clauses. By combining ORDER BY and LIMIT, its possible to get the top X entries by some criteria, and
then bring in additional data from the graph.
Another use is to filter on aggregated values. WITH is used to introduce aggregates which can then by
used in predicates in WHERE. These aggregate expressions create new bindings in the results. WITH can
also, just like RETURN, alias expressions that are introduced into the results using the aliases as binding
name.
is also used to separate reading from updating of the graph. Every part of a query must be either
read-only or write-only. When going from a writing part to a reading part, the switch must be done
with a WITH clause.
WITH
Figure9.5.Graph
nam e = 'David'
KNOWS
nam e = 'Anders'
BLOCKS
BLOCKS
KNOWS
nam e = 'Ceasar'
nam e = 'Bossm an'
KNOWS
KNOWS
nam e = 'Em il'
9.5.1.Filter on aggregate function results

Aggregated results have to pass through a WITH clause to be able to filter on.
Query.
MATCH (david { name: "David" })--(otherPerson)-->()
WITH otherPerson, count(*) AS foaf
WHERE foaf > 1
RETURN otherPerson
The person connected to David with the at least more than one outgoing relationship will be returned
by the query.
123
General Clauses
Result
otherPerson
Node[2]{name:"Anders"}
1 row
9.5.2.Sort results before using collect on them

You can sort your results before passing them to collect, thus sorting the resulting collection.
Query.
MATCH (n)
WITH n
ORDER BY n.name DESC LIMIT 3
RETURN collect(n.name)
A list of the names of people in reverse order, limited to 3, in a collection.

Result
collect(n.name)
["Emil", "David", "Ceasar"]
1 row
9.5.3.Limit branching of your path search

You can match paths, limit to a certain number, and then match again using those paths as a base As
well as any number of similar limited searches.
Query.
MATCH (n { name: "Anders" })--(m)
WITH m
ORDER BY m.name DESC LIMIT 1
MATCH (m)--(o)
RETURN o.name
Starting at Anders, find all matching nodes, order by name descending and get the top result, then find
all the nodes connected to that top result, and return their names.
Result
o.name
"Anders"
"Bossman"
2 rows
124
General Clauses
9.6.Union
Combining results from multiple queries is done through the UNION operator.
Combines the results of two or more queries into a single result set that includes all the rows that
belong to all queries in the union.
The number and the names of the columns must be identical in all queries combined by using UNION.
To keep all the result rows, use UNION ALL. Using just UNION will combine and remove duplicates from
the result set.
Figure9.6.Graph
Act or
Act or
nam e = 'Ant hony Hopkins'
nam e = 'Hit chcock'
KNOWS
Act or
ACTS_IN
nam e = 'Helen Mirren'
ACTS_IN
Movie
t it le = 'Hit chcock'
9.6.1.Combine two queries

Combining the results from two queries is done using UNION ALL.
Query.
MATCH (n:Actor)
RETURN n.name AS name
UNION ALL MATCH (n:Movie)
RETURN n.title AS name
The combined result is returned, including duplicates.

Result
name
"Anthony Hopkins"
"Helen Mirren"
"Hitchcock"
"Hitchcock"
4 rows
9.6.2.Combine two queries and remove duplicates

By not including ALL in the UNION, duplicates are removed from the combined result set
125
General Clauses
Query.
MATCH (n:Actor)
RETURN n.name AS name
UNION
MATCH (n:Movie)
RETURN n.title AS name
The combined result is returned, without duplicates.

Result
name
"Anthony Hopkins"
"Helen Mirren"
"Hitchcock"
3 rows
126
General Clauses
9.7.Using
If you do not specify an explicit START clause, Cypher needs to infer where in the graph to start your
query. This is done by looking at the WHERE clause and the MATCH clause and using that information to
find a useful index.
This index might not be the best choice though sometimes multiple indexes could be used, and
Cypher has picked the wrong one (from a performance point of view).
You can force Cypher to use a specific starting point by using the USING clause. This is called giving
Cypher an index hint.
If your query matches large parts of an index, it might be faster to scan the label and filter out nodes
that do not match. To do this, you can use USING SCAN. It will force Cypher to not use an index that
could have been used, and instead do a label scan.
Note
You cannot use index hints if your query has a START clause.
9.7.1.Query using an index hint

To query using an index hint, use USING INDEX.
Query.
MATCH (n:Swedish)
USING INDEX n:Swedish(surname)
WHERE n.surname = 'Taylor'
RETURN n
The query result is returned as usual.

Result
n
Node[3]{name:"Andres", age:36, awesome:true, surname:"Taylor"}
1 row
9.7.2.Query using multiple index hints

To query using multiple index hints, use USING INDEX.
Query.
MATCH (m:German)-->(n:Swedish)
USING INDEX m:German(surname)
USING INDEX n:Swedish(surname)
WHERE m.surname = 'Plantikow' AND n.surname = 'Taylor'
RETURN m
The query result is returned as usual.

Result
m
Node[1]{name:"Stefan", surname:"Plantikow"}
1 row
127
General Clauses
9.7.3.Hinting a label scan

If the best performance is to be had by scanning all nodes in a label and then filtering on that set, use
USING SCAN.
Query.
MATCH (m:German)
USING SCAN m:German
WHERE m.surname = 'Plantikow'
RETURN m
This query does its work by finding all :German labeled nodes and filtering them by the surname
property.
Result
m
Node[1]{name:"Stefan", surname:"Plantikow"}
1 row
128
Chapter10.Reading Clauses
The flow of data within a Cypher query is an unordered sequence of maps with key-value pairs a
set of possible bindings between the identifiers in the query and values derived from the database.
This set is refined and augmented by subsequent parts of the query.
129
Reading Clauses
10.1.Match
10.1.1.Introduction
MATCH
is the primary way of getting data from the database into the current set of bindings.
The MATCH clause allows you to specify the patterns Cypher will search for in the database.
Nodes and relationships that are already known at this stage are called bound pattern elements.
Cypher will now try to find the unknown parts of the pattern.
If MATCH is the first clause in your query, nothing is bound at this stage. Cypher needs starting points to
do its pattern matching.
If no bound nodes exist, Cypher can scan all nodes in the database, all nodes with a certain label, or
it can use indexes to quickly find the relevant starting points. For more information on indexes, see
Section13.1, Indexes. If you want to use index hints to force Cypher to use a specific index, read
more in Section9.7, Using.
defines the MATCH patterns in more detail. The predicates are part of the pattern description, not a
filter applied after the matching is done. This means that WHERE should always be put together with the
MATCH clause it belongs to.
WHERE
After finding starting points either by using scans, indexes or already bound points the execution
engine will use pattern matching to find matching subgraphs. As Cypher is declarative, it can change
the order of these operations. Predicates in WHERE clauses can be evaluated before pattern matching,
during pattern matching, or after finding matches.
Tip
To understand the patterns used in the MATCH clause, read Section8.5, Patterns.
The following graph is used for the examples below:
Figure10.1.Graph
Person
nam e = 'Charlie Sheen'
FATHER
Person
Person
Person
Person
ACTED_IN
nam e = 'Oliver St one'
nam e = 'Michael Douglas'
DIRECTED
ACTED_IN
ACTED_IN
Movie
nam e = 'Mart in Sheen'
ACTED_IN
ACTED_IN
nam e = 'Rob Reiner'
DIRECTED
Movie
nam e = 'WallSt reet '

t it le = 'Wall St reet '
nam e = 'TheAm ericanPresident '

t it le = 'The Am erican President '
10.1.2.Basic node finding

Get all nodes
By just specifying a pattern with a single node and no labels, all nodes in the graph will be returned.
Query.
MATCH (n)
130
Reading Clauses
RETURN n
Returns all the nodes in the database.

Result
n
Node[0]{name:"Oliver Stone"}
Node[1]{name:"Charlie Sheen"}
Node[2]{name:"Martin Sheen"}
Node[3]{name:"TheAmericanPresident", title:"The American President"}
Node[4]{name:"WallStreet", title:"Wall Street"}
Node[5]{name:"Rob Reiner"}
Node[6]{name:"Michael Douglas"}
7 rows
Get all nodes with a label
Getting all nodes with a label on them is done with a single node pattern where the node has a label on
it.
Query.
MATCH (movie:Movie)
RETURN movie
Returns all the movies in the database.

Result
movie
2 rows
Related nodes
The symbol -- means related to, without regard to type or direction of the relationship.
Query.
MATCH (director { name:'Oliver Stone' })--(movie)
RETURN movie.title
Returns all the movies directed by Oliver Stone.

Result
movie.title
"Wall Street"
1 row
Match with labels
To constrain your pattern with labels on nodes, you add it to your pattern nodes, using the label
syntax.
131
Reading Clauses
Query.
MATCH (charlie:Person { name:'Charlie Sheen' })--(movie:Movie)
RETURN movie
Return any nodes connected with the Person Charlie that are labeled Movie.
Result
movie
1 row
10.1.3.Relationship basics
Outgoing relationships
When the direction of a relationship is interesting, it is shown by using --> or <--, like this:
Query.
MATCH (martin { name:'Martin Sheen' })-->(movie)
RETURN movie.title
Returns nodes connected to Martin by outgoing relationships.

Result
movie.title
"Wall Street"
"The American President"
2 rows
Directed relationships and identifier
If an identifier is needed, either for filtering on properties of the relationship, or to return the
relationship, this is how you introduce the identifier.
Query.
MATCH (martin { name:'Martin Sheen' })-[r]->(movie)
RETURN r
Returns all outgoing relationships from Martin.

Result
r
:ACTED_IN[1]{}
:ACTED_IN[3]{}
2 rows
Match by relationship type
When you know the relationship type you want to match on, you can specify it by using a colon
together with the relationship type.
Query.
132
Reading Clauses
MATCH (wallstreet { title:'Wall Street' })<-[:ACTED_IN]-(actor)
RETURN actor
Returns nodes that ACTED_IN Wall Street.

Result
actor
3 rows
Match by multiple relationship types
To match on one of multiple types, you can specify this by chaining them together with the pipe
symbol |.
Query.
MATCH (wallstreet { title:'Wall Street' })<-[:ACTED_IN|:DIRECTED]-(person)
RETURN person
Returns nodes with a ACTED_IN or DIRECTED relationship to Wall Street.

Result
person
4 rows
Match by relationship type and use an identifier
If you both want to introduce an identifier to hold the relationship, and specify the relationship type
you want, just add them both, like this.
Query.
MATCH (wallstreet { title:'Wall Street' })<-[r:ACTED_IN]-(actor)
RETURN r
Returns nodes that ACTED_IN Wall Street.

Result
r
:ACTED_IN[0]{}
:ACTED_IN[1]{}
:ACTED_IN[2]{}
3 rows
133
Reading Clauses
10.1.4.Relationships in depth
Relationship types with uncommon characters
Sometime your database will have types with non-letter characters, or with spaces in them. Use `
(backtick) to quote these.
Query.
MATCH (n { name:'Rob Reiner' })-[r:`TYPE THAT HAS SPACE IN IT`]->()
RETURN r
Returns a relationship of a type with spaces in it.

Result
r
:TYPE THAT HAS SPACE IN IT[8]{}
1 row
Multiple relationships
Relationships can be expressed by using multiple statements in the form of ()--(), or they can be
strung together, like this:
Query.
MATCH (charlie { name:'Charlie Sheen' })-[:ACTED_IN]->(movie)<-[:DIRECTED]->(director)
RETURN charlie,movie,director
Returns the three nodes in the path.

Result
charlie
movie
director
Node[4]{name:"WallStreet",
title:"Wall Street"}
1 row
Variable length relationships
Nodes that are a variable number of relationshipnode hops away can be found using the following
syntax: -[:TYPE*minHops..maxHops]->. minHops and maxHops are optional and default to 1 and infinity
respectively. When no bounds are given the dots may be omitted.
Query.
MATCH (martin { name:"Martin Sheen" })-[:ACTED_IN*1..2]-(x)
RETURN x
Returns nodes that are 1 or 2 relationships away from Martin.

Result
x
5 rows
134
Reading Clauses
x
5 rows
Relationship identifier in variable length relationships
When the connection between two nodes is of variable length, a relationship identifier becomes an
collection of relationships.
Query.
MATCH (actor { name:'Charlie Sheen' })-[r:ACTED_IN*2]-(co_actor)
RETURN r
The query returns a collection of relationships.

Result
r
[:ACTED_IN[0]{}, :ACTED_IN[1]{}]
[:ACTED_IN[0]{}, :ACTED_IN[2]{}]
2 rows
Match with properties on a variable length path
A variable length relationship with properties defined on in it means that all relationships in the path
must have the property set to the given value. In this query, there are two paths between Charile Sheen
and his dad Martin Sheen. One of the includes a blocked relationship and the other doesnt. In this
case we first alter the original graph by using the following query to add blocked and unblocked
relationships:
MATCH (charlie:Person { name:'Charlie Sheen' }),(martin:Person { name:'Martin Sheen' })
CREATE (charlie)-[:X { blocked:false }]->(:Unblocked)<-[:X { blocked:false }]-(martin)
CREATE (charlie)-[:X { blocked:true }]->(:Blocked)<-[:X { blocked:false }]-(martin);
This means that we are starting out with the following graph:
Person
FATHER
Person
Person
ACTED_IN
DIRECTED
ACTED_IN
Movie
X
blocked = t rue
ACTED_IN
Person
Person
ACTED_IN
Blocked

X
blocked = false
ACTED_IN
Movie
X
blocked = false
X
blocked = false
DIRECTED
Unblocked

Query.
MATCH p =(charlie:Person)-[* { blocked:false }]-(martin:Person)
WHERE charlie.name = 'Charlie Sheen' AND martin.name = 'Martin Sheen'
RETURN p
Returns the paths between Charlie and Martin Sheen where all relationships have the blocked property
set to FALSE.
135
Reading Clauses
Result
p
[Node[1]{name:"Charlie Sheen"}, :X[12]{blocked:false}, Node[9]{}, :X[13]{blocked:false}, Node[2]
{name:"Martin Sheen"}]
1 row
Zero length paths
Using variable length paths that have the lower bound zero means that two identifiers can point
to the same node. If the distance between two nodes is zero, they are by definition the same node.
Note that when matching zero length paths the result may contain a match even when matching on a
relationship type not in use.
Query.
MATCH (wallstreet:Movie { title:'Wall Street' })-[*0..1]-(x)
RETURN x
Returns all nodes that are zero or one relationships away from Wall Street.
Result
x
5 rows
Named path
If you want to return or filter on a path in your pattern graph, you can a introduce a named path.
Query.
MATCH p =(michael { name:'Michael Douglas' })-->()
RETURN p
Returns the two paths starting from Michael.

Result
p
[Node[6]{name:"Michael Douglas"}, :ACTED_IN[2]{}, Node[4]{name:"WallStreet", title:"Wall Street"}]
[Node[6]{name:"Michael Douglas"}, :ACTED_IN[4]{}, Node[3]{name:"TheAmericanPresident", title:"The
American President"}]
2 rows
Matching on a bound relationship
When your pattern contains a bound relationship, and that relationship pattern doesnt specify
direction, Cypher will try to match the relationship in both directions.
136
Reading Clauses
Query.
MATCH (a)-[r]-(b)
WHERE id(r)= 0
RETURN a,b
This returns the two connected nodes, once as the start node, and once as the end node.
Result
a
2 rows
10.1.5.Shortest path
Single shortest path
Finding a single shortest path between two nodes is as easy as using the shortestPath function. Its
done like this:
Query.
MATCH (martin:Person { name:"Martin Sheen" }),(oliver:Person { name:"Oliver Stone" }),
p = shortestPath((martin)-[*..15]-(oliver))
RETURN p
This means: find a single shortest path between two nodes, as long as the path is max 15 relationships
long. Inside of the parentheses you define a single link of a path the starting node, the connecting
relationship and the end node. Characteristics describing the relationship like relationship type, max
hops and direction are all used when finding the shortest path. You can also mark the path as optional.
Result
p
[Node[2]{name:"Martin Sheen"}, :ACTED_IN[1]{}, Node[4]{name:"WallStreet", title:"Wall
Street"}, :DIRECTED[5]{}, Node[0]{name:"Oliver Stone"}]
1 row
All shortest paths
Finds all the shortest paths between two nodes.
Query.
MATCH (martin:Person { name:"Martin Sheen" }),(michael:Person { name:"Michael Douglas" }),
p = allShortestPaths((martin)-[*]-(michael))
RETURN p
Finds the two shortest paths between Martin and Michael.

Result
p
[Node[2]{name:"Martin Sheen"}, :ACTED_IN[3]{}, Node[3]{name:"TheAmericanPresident", title:"The
American President"}, :ACTED_IN[4]{}, Node[6]{name:"Michael Douglas"}]
2 rows
137
Reading Clauses
p
[Node[2]{name:"Martin Sheen"}, :ACTED_IN[1]{}, Node[4]{name:"WallStreet", title:"Wall
Street"}, :ACTED_IN[2]{}, Node[6]{name:"Michael Douglas"}]
2 rows
138
Reading Clauses
10.2.Optional Match
10.2.1.Introduction
matches patterns against your graph database, just like MATCH does. The difference is
that if no matches are found, OPTIONAL MATCH will use NULLs for missing parts of the pattern. OPTIONAL
MATCH could be considered the Cypher equivalent of the outer join in SQL.
OPTIONAL MATCH
Either the whole pattern is matched, or nothing is matched. Remember that WHERE is part of the pattern
description, and the predicates will be considered while looking for matches, not after. This matters
especially in the case of multiple (OPTIONAL) MATCH clauses, where it is crucial to put WHERE together with
the MATCH it belongs to.
Tip
To understand the patterns used in the OPTIONAL MATCH clause, read Section8.5, Patterns.
Figure10.2.Graph
Person
FATHER
Person
Person
Person
Person
ACTED_IN
DIRECTED
ACTED_IN
ACTED_IN
Movie
ACTED_IN
ACTED_IN
DIRECTED
Movie


10.2.2.Relationship
If a relationship is optional, use the OPTIONAL MATCH clause. This is similar to how a SQL outer join
works. If the relationship is there, it is returned. If its not, NULL is returned in its place.
Query.
MATCH (a:Movie { title: 'Wall Street' })
OPTIONAL MATCH (a)-->(x)
RETURN x
Returns NULL, since the node has no outgoing relationships.

Result
x
<null>
1 row
10.2.3.Properties on optional elements

Returning a property from an optional element that is NULL will also return NULL.
139
Reading Clauses
Query.
OPTIONAL MATCH (a)-->(x)
RETURN x, x.name
Returns the element x (NULL in this query), and NULL as its name.
Result
x
x.name
<null>
<null>
1 row
10.2.4.Optional typed and named relationship

Just as with a normal relationship, you can decide which identifier it goes into, and what relationship
type you need.
Query.
OPTIONAL MATCH (a)-[r:ACTS_IN]->()
RETURN r
This returns a node, and NULL, since the node has no outgoing ACTS_IN relationships.
Result
r
<null>
1 row
140
Reading Clauses
10.3.Where
WHERE
is not a clause in its own right rather, its part of MATCH, OPTIONAL MATCH, START and WITH.
In the case of WITH and START, WHERE simply filters the results.
For MATCH and OPTIONAL MATCH on the other hand, WHERE adds constraints to the patterns described. It
should not be seen as a filter after the matching is finished.
Note
In the case of multiple (OPTIONAL) MATCH clauses, the predicate in WHERE is always a part of
the patterns in the directly preceding MATCH. Both results and performance may be impacted
if the WHERE is put inside the wrong MATCH clause.
Figure10.3.Graph
Swedish
nam e = 'Andres'
age = 36
belt = 'whit e'
KNOWS
nam e = 'Tobias'
age = 25
KNOWS
nam e = 'Pet er'

age = 34
10.3.1.Basic usage
Boolean operations
You can use the expected boolean operators AND and OR, and also the boolean function NOT. See
Section8.7, Working with NULL for more information on how this works with NULL.
Query.
MATCH (n)
WHERE n.name = 'Peter' XOR (n.age < 30 AND n.name = "Tobias") OR NOT (n.name = "Tobias" OR
n.name="Peter")
RETURN n
This query shows how boolean operators can be used.

Result
n
Node[0]{name:"Tobias", age:25}
Node[1]{name:"Peter", age:34}
Node[2]{name:"Andres", age:36, belt:"white"}
3 rows
Filter on node label
To filter nodes by label, write a label predicate after the WHERE keyword using WHERE n:foo.
Query.
141
Reading Clauses
MATCH (n)
WHERE n:Swedish
RETURN n
The "Andres" node will be returned.

Result
n
1 row
Filter on node property
To filter on a property, write your clause after the WHERE keyword. Filtering on relationship properties
works just the same way.
Query.
MATCH (n)
WHERE n.age < 30
RETURN n
The "Tobias" node will be returned.

Result
n
1 row
Property exists
To only include nodes/relationships that have a property, use the HAS() function and just write out the
identifier and the property you expect it to have.
Query.
MATCH (n)
WHERE HAS (n.belt)
RETURN n
The node named "Andres" is returned.

Result
n
1 row
10.3.2.Regular expressions
Regular expressions
You can match on regular expressions by using =~ "regexp", like this:
Query.
MATCH (n)
WHERE n.name =~ 'Tob.*'
142
Reading Clauses
RETURN n
The "Tobias" node will be returned.

Result
n
1 row
Escaping in regular expressions
If you need a forward slash inside of your regular expression, escape it. Remember that back slash
needs to be escaped in string literals
Query.
MATCH (n)
WHERE n.name =~ 'Some\/thing'
RETURN n
No nodes match this regular expression.

Result
n
(empty result)
0 row
Case insensitive regular expressions
By pre-pending a regular expression with (?i), the whole expression becomes case insensitive.
Query.
MATCH (n)
WHERE n.name =~ '(?i)ANDR.*'
RETURN n
The node with name "Andres" is returned.

Result
n
1 row
10.3.3.Using patterns in WHERE

Filter on patterns
Patterns are expressions in Cypher, expressions that return a collection of paths. Collection
expressions are also predicates an empty collection represents false, and a non-empty represents
true.
So, patterns are not only expressions, they are also predicates. The only limitation to your pattern is
that you must be able to express it in a single path. You can not use commas between multiple paths
like you do in MATCH. You can achieve the same effect by combining multiple patterns with AND.
143
Reading Clauses
Note that you can not introduce new identifiers here. Although it might look very similar to the MATCH
patterns, the WHERE clause is all about eliminating matched subgraphs. MATCH (a)-[*]->(b) is very
different from WHERE (a)-[*]->(b); the first will produce a subgraph for every path it can find between
a and b, and the latter will eliminate any matched subgraphs where a and b do not have a directed
relationship chain between them.
Query.
MATCH (tobias { name: 'Tobias' }),(others)
WHERE others.name IN ['Andres', 'Peter'] AND (tobias)<--(others)
RETURN others
Nodes that have an outgoing relationship to the "Tobias" node are returned.
Result
others
1 row
Filter on patterns using NOT
The NOT function can be used to exclude a pattern.
Query.
MATCH (persons),(peter { name: 'Peter' })
WHERE NOT (persons)-->(peter)
RETURN persons
Nodes that do not have an outgoing relationship to the "Peter" node are returned.
Result
persons
2 rows
Filter on patterns with properties
You can also add properties to your patterns:
Query.
MATCH (n)
WHERE (n)-[:KNOWS]-({ name:'Tobias' })
RETURN n
Finds all nodes that have a KNOWS relationship to a node with the name Tobias.
Result
n
1 row
144
Reading Clauses
Filtering on relationship type
You can put the exact relationship type in the MATCH pattern, but sometimes you want to be able to do
more advanced filtering on the type. You can use the special property TYPE to compare the type with
something else. In this example, the query does a regular expression comparison with the name of the
relationship type.
Query.
MATCH (n)-[r]->()
WHERE n.name='Andres' AND type(r)=~ 'K.*'
RETURN r
This returns relationships that has a type whose name starts with K.
Result
r
:KNOWS[0]{}
:KNOWS[1]{}
2 rows
10.3.4.Collections
IN operator
To check if an element exists in a collection, you can use the IN operator.
Query.
MATCH (a)
WHERE a.name IN ["Peter", "Tobias"]
RETURN a
This query shows how to check if a property exists in a literal collection.

Result
a
2 rows
10.3.5.Missing properties and values

Default to false if property is missing
As missing properties evaluate to NULL, the comparision in the example will evaluate to FALSE for nodes
without the belt property.
Query.
MATCH (n)
WHERE n.belt = 'white'
RETURN n
Only nodes with the belt property are returned.

145
Reading Clauses
Result
n
1 row
Default to true if property is missing
If you want to compare a property on a graph element, but only if it exists, you can compare the
property against both the value you are looking for and NULL, like:
Query.
MATCH (n)
WHERE n.belt = 'white' OR n.belt IS NULL RETURN n
ORDER BY n.name
This returns all nodes, even those without the belt property.
Result
n
3 rows
Filter on NULL
Sometimes you might want to test if a value or an identifier is NULL. This is done just like SQL does it,
with IS NULL. Also like SQL, the negative is IS NOT NULL, although NOT(IS NULL x) also works.
Query.
MATCH (person)
WHERE person.name = 'Peter' AND person.belt IS NULL RETURN person
Nodes that Tobias is not connected to are returned.

Result
person
1 row
146
Reading Clauses
10.4.Start
Every query describes a pattern, and in that pattern one can have multiple starting points. A starting
point is a relationship or a node where a pattern is anchored. You can either introduce starting points
by id, or by index lookups. Note that trying to use an index that doesnt exist will generate an error.
Note
is optional. If you do not specify explicit starting points, Cypher will try and infer
starting points from your query. This is done based on node labels and predicates contained
in your query. See Chapter13, Schema for more information. In general, the START clause is
only really needed when using legacy indexes.
START
This is the graph the examples are using:

Figure10.4.Graph
Node[ 0] nam e = 'A'
KNOWS
Node[ 1] nam e = 'B'
KNOWS
Node[ 2] nam e = 'C'
10.4.1.Get node or relationship from index

Node by index lookup
When the starting point can be found by using index lookups, it can be done like this: node:indexname(key = "value"). In this example, there exists a node index named nodes.
Query.
START n=node:nodes(name = "A")
RETURN n
The query returns the node indexed with the name "A".
Result
n
Node[0]{name:"A"}
1 row
Relationship by index lookup
When the starting point can be found by using index lookups, it can be done like this:
relationship:index-name(key = "value").
Query.
START r=relationship:rels(name = "Andrs")
RETURN r
The relationship indexed with the name property set to "Andrs" is returned by the query.
147
Reading Clauses
Result
r
:KNOWS[0]{name:"Andrs"
1 row
Node by index query
When the starting point can be found by more complex Lucene queries, this is the syntax to use:
node:index-name("query").This allows you to write more advanced index queries.
Query.
START n=node:nodes("name:A")
RETURN n
The node indexed with name "A" is returned by the query.

Result
n
Node[0]{name:"A"}
1 row
10.4.2.Get node or relationship by id

Node by id
Binding a node as a starting point is done with the node(*) function.
Note
Neo4j reuses its internal ids when nodes and relationships are deleted, which means its
bad practice to refer to them this way. Instead, use application generated ids.
Query.
START n=node(0)
RETURN n
The corresponding node is returned.

Result
n
Node[0]{name:"A"}
1 row
Relationship by id
Binding a relationship as a starting point is done with the relationship(*) function, which can also be
abbreviated rel(*). See the section called Node by id for more information on Neo4j ids.
Query.
START r=relationship(0)
RETURN r
The relationship with id 0 is returned.

148
Reading Clauses
Result
r
:KNOWS[0]{}
1 row
Multiple nodes by id
Multiple nodes are selected by listing them separated by commas.
Query.
START n=node(0, 1, 2)
RETURN n
This returns the nodes listed in the START statement.

Result
n
Node[0]{name:"A"}
Node[1]{name:"B"}
Node[2]{name:"C"}
3 rows
10.4.3.Get multiple or all nodes

All nodes
To get all the nodes, use an asterisk. This can be done with relationships as well.
Tip
The preferred way to do this is to use a MATCH clause, see the section called Get all nodes
in Section10.1, Match for how to do that.
Query.
START n=node(*)
RETURN n
This query returns all the nodes in the graph.

Result
n
Node[0]{name:"A"}
Node[1]{name:"B"}
Node[2]{name:"C"}
3 rows
Multiple starting points
Sometimes you want to bind multiple starting points. Just list them separated by commas.
Query.
149
Reading Clauses
START a=node(0), b=node(1)
RETURN a,b
Both the nodes A and the B are returned.

Result
a
Node[0]{name:"A"}
Node[1]{name:"B"}
1 row
150
Reading Clauses
10.5.Aggregation
10.5.1.Introduction
To calculate aggregated data, Cypher offers aggregation, much like SQLs GROUP BY.
Aggregate functions take multiple input values and calculate an aggregated value from them.
Examples are avg that calculates the average of multiple numeric values, or min that finds the smallest
numeric value in a set of values.
Aggregation can be done over all the matching sub graphs, or it can be further divided by introducing
key values. These are non-aggregate expressions, that are used to group the values going into the
aggregate functions.
So, if the return statement looks something like this:
RETURN n, count(*)
We have two return expressions n, and count(*). The first, n, is no aggregate function, and so it will
be the grouping key. The latter, count(*) is an aggregate expression. So the matching subgraphs will
be divided into different buckets, depending on the grouping key. The aggregate function will then run
on these buckets, calculating the aggregate values.
If you want to use aggregations to sort your result set, the aggregation must be included in the RETURN
to be used in your ORDER BY.
The last piece of the puzzle is the DISTINCT keyword. It is used to make all values unique before
running them through an aggregate function.
An example might be helpful. In this case, we are running the query against the following data:
Person
nam e = 'A'
propert y = 13
KNOWS
Person
nam e = 'D'
eyes = 'brown'
KNOWS
KNOWS
Person
Person
nam e = 'B'
propert y = 33
eyes = 'blue'
nam e = 'C'
propert y = 44
eyes = 'blue'
KNOWS
Person
nam e = 'D'
Query.
MATCH (me:Person)-->(friend:Person)-->(friend_of_friend:Person)
WHERE me.name = 'A'
RETURN count(DISTINCT friend_of_friend), count(friend_of_friend)
151
KNOWS
Reading Clauses
In this example we are trying to find all our friends of friends, and count them. The first aggregate
function, count(DISTINCT friend_of_friend), will only see a friend_of_friend once DISTINCT removes
the duplicates. The latter aggregate function, count(friend_of_friend), might very well see the same
friend_of_friend multiple times. In this case, both B and C know D and thus D will get counted twice,
when not using DISTINCT.
Result
count(distinct friend_of_friend)
count(friend_of_friend)
1 row
The following examples are assuming the example graph structure below.
Figure10.5.Graph
Person
nam e = 'A'
propert y = 13
KNOWS
Person
nam e = 'D'
eyes = 'brown'
KNOWS
KNOWS
Person
Person
nam e = 'B'
propert y = 33
eyes = 'blue'
nam e = 'C'
propert y = 44
eyes = 'blue'
10.5.2.COUNT
is used to count the number of rows. COUNT can be used in two forms COUNT(*) which just
counts the number of matching rows, and COUNT(<identifier>), which counts the number of non-NULL
values in <identifier>.
COUNT
Count nodes
To count the number of nodes, for example the number of nodes connected to one node, you can use
count(*).
Query.
MATCH (n { name: 'A' })-->(x)
RETURN n, count(*)
This returns the start node and the count of related nodes.
Result
n
count(*)
Node[1]{name:"A", property:13}
1 row
Group Count Relationship Types
To count the groups of relationship types, return the types and count them with count(*).
152
Reading Clauses
Query.
MATCH (n { name: 'A' })-[r]->()
RETURN type(r), count(*)
The relationship types and their group count is returned by the query.
Result
type(r)
count(*)
"KNOWS"
1 row
Count entities
Instead of counting the number of results with count(*), it might be more expressive to include the
name of the identifier you care about.
Query.
MATCH (n { name: 'A' })-->(x)
RETURN count(x)
The example query returns the number of connected nodes from the start node.
Result
count(x)
3
1 row
Count non-null values
You can count the non-null values by using count(<identifier>).
Query.
MATCH (n:Person)
RETURN count(n.property)
The count of related nodes with the property property set is returned by the query.
Result
count(n.property)
3
1 row
10.5.3.Statistics
sum
The sum aggregation function simply sums all the numeric values it encounters. NULLs are silently
dropped.
Query.
MATCH (n:Person)
RETURN sum(n.property)
153
Reading Clauses
This returns the sum of all the values in the property property.
Result
sum(n.property)
90
1 row
avg
avg
calculates the average of a numeric column.
Query.
MATCH (n:Person)
RETURN avg(n.property)
The average of all the values in the property property is returned by the example query.
Result
avg(n.property)
30. 0
1 row
percentileDisc
calculates the percentile of a given value over a group, with a percentile from 0.0 to
1.0. It uses a rounding method, returning the nearest value to the percentile. For interpolated values,
see percentileCont.
percentileDisc
Query.
MATCH (n:Person)
RETURN percentileDisc(n.property, 0.5)
The 50th percentile of the values in the property property is returned by the example query. In this
case, 0.5 is the median, or 50th percentile.
Result
percentileDisc(n.property, 0.5)
33
1 row
percentileCont
calculates the percentile of a given value over a group, with a percentile from 0.0 to
1.0. It uses a linear interpolation method, calculating a weighted average between two values, if the
desired percentile lies between them. For nearest values using a rounding method, see percentileDisc.
percentileCont
Query.
MATCH (n:Person)
RETURN percentileCont(n.property, 0.4)
The 40th percentile of the values in the property property is returned by the example query, calculated
with a weighted average.
154
Reading Clauses
Result
percentileCont(n.property, 0.4)
29. 0
1 row
stdev
calculates the standard deviation for a given value over a group. It uses a standard two-pass
method, with N - 1 as the denominator, and should be used when taking a sample of the population for
an unbiased estimate. When the standard variation of the entire population is being calculated, stdevp
should be used.
stdev
Query.
MATCH (n)
WHERE n.name IN ['A', 'B', 'C']
RETURN stdev(n.property)
The standard deviation of the values in the property property is returned by the example query.
Result
stdev(n.property)
15. 716233645501712
1 row
stdevp
calculates the standard deviation for a given value over a group. It uses a standard two-pass
method, with N as the denominator, and should be used when calculating the standard deviation for an
entire population. When the standard variation of only a sample of the population is being calculated,
stdev should be used.
stdevp
Query.
MATCH (n)
WHERE n.name IN ['A', 'B', 'C']
RETURN stdevp(n.property)
The population standard deviation of the values in the property property is returned by the example
query.
Result
stdevp(n.property)
12. 832251036613439
1 row
max
max
find the largest value in a numeric column.
Query.
MATCH (n:Person)
RETURN max(n.property)
The largest of all the values in the property property is returned.

155
Reading Clauses
Result
max(n.property)
44
1 row
min
min
takes a numeric property as input, and returns the smallest value in that column.
Query.
MATCH (n:Person)
RETURN min(n.property)
This returns the smallest of all the values in the property property.
Result
min(n.property)
13
1 row
10.5.4.collect
collect
collects all the values into a list. It will ignore NULLs.
Query.
MATCH (n:Person)
RETURN collect(n.property)
Returns a single row, with all the values collected.

Result
collect(n.property)
[13, 33, 44]
1 row
10.5.5.DISTINCT
All aggregation functions also take the DISTINCT modifier, which removes duplicates from the values.
So, to count the number of unique eye colors from nodes related to a, this query can be used:
Query.
MATCH (a:Person { name: 'A' })-->(b)
RETURN count(DISTINCT b.eyes)
Returns the number of eye colors.

Result
count(distinct b.eyes)
2
1 row
156
Chapter11.Writing Clauses
Write data to the database.
157
Writing Clauses
11.1.Create
Creating graph elements nodes and relationships, is done with CREATE.
Tip
In the CREATE clause, patterns are used a lot. Read Section8.5, Patterns for an
introduction.
11.1.1.Create nodes
Create single node
Creating a single node is done by issuing the following query.
Query.
CREATE (n)
Nothing is returned from this query, except the count of affected nodes.
Result
(empty result)
Nodes created: 1
Create a node with a label
To add a label when creating a node, use the syntax below.
Query.
CREATE (n:Person)
Nothing is returned from this query.

Result
(empty result)
Nodes created: 1
Labels added: 1
Create a node with multiple labels
To add labels when creating a node, use the syntax below. In this case, we add two labels.
Query.
CREATE (n:Person:Swedish)

Result
(empty result)
Nodes created: 1
Labels added: 2
Create node and add labels and properties
When creating a new node with labels, you can add properties at the same time.
158
Writing Clauses
Query.
CREATE (n:Person { name : 'Andres', title : 'Developer' })

Result
(empty result)
Nodes created: 1
Properties set: 2
Labels added: 1
Return created node
Creating a single node is done by issuing the following query.
Query.
CREATE (a { name : 'Andres' })
RETURN a
The newly created node is returned.

Result
a
Node[1]{name:"Andres"}
1 row
Nodes created: 1
Properties set: 1
11.1.2.Create relationships
Create a relationship between two nodes
To create a relationship between two nodes, we first get the two nodes. Once the nodes are loaded, we
simply create a relationship between them.
Query.
MATCH (a:Person),(b:Person)
WHERE a.name = 'Node A' AND b.name = 'Node B'
CREATE (a)-[r:RELTYPE]->(b)
RETURN r
The created relationship is returned by the query.

Result
r
:RELTYPE[1]{}
1 row
159
Writing Clauses
Create a relationship and set properties
Setting properties on relationships is done in a similar manner to how its done when creating nodes.
Note that the values can be any expression.
Query.
MATCH (a:Person),(b:Person)
WHERE a.name = 'Node A' AND b.name = 'Node B'
CREATE (a)-[r:RELTYPE { name : a.name + '<->' + b.name }]->(b)
RETURN r
The newly created relationship is returned by the example query.

Result
r
:RELTYPE[1]{name:"Node A<->Node B"}
1 row
Properties set: 1
11.1.3.Create a full path

When you use CREATE and a pattern, all parts of the pattern that are not already in scope at this time
will be created.
Query.
CREATE p =(andres { name:'Andres' })-[:WORKS_AT]->(neo)<-[:WORKS_AT]-(michael { name:'Michael' })
RETURN p
This query creates three nodes and two relationships in one go, assigns it to a path identifier, and
returns it.
Result
p
[Node[3]{name:"Andres"}, :WORKS_AT[2]{}, Node[4]{}, :WORKS_AT[3]{}, Node[5]{name:"Michael"}]
1 row
Nodes created: 3
Properties set: 2
11.1.4.Use parameters with CREATE

Create node with a parameter for the properties
You can also create a graph entity from a map. All the key/value pairs in the map will be set as
properties on the created relationship or node. In this case we add a Person label to the node as well.
Parameters.
{
"props" : {
"name" : "Andres",
160
Writing Clauses
"position" : "Developer"
}
}
Query.
RETURN n
Result
n
Node[1]{name:"Andres", position:"Developer"}
1 row
Nodes created: 1
Properties set: 2
Labels added: 1
Create multiple nodes with a parameter for their properties
By providing Cypher an array of maps, it will create a node for each map.
Note
When you do this, you cant create anything else in the same CREATE clause.
Parameters.
{
"props" : [ {
"name" : "Andres",
}, {
"name" : "Michael",
} ]
}
Query.
CREATE (n { props })
RETURN n
Result
n
Node[2]{name:"Andres", position:"Developer"}
Node[3]{name:"Michael", position:"Developer"}
2 rows
Nodes created: 2
Properties set: 4
161
Writing Clauses
11.2.Create Unique
CREATE UNIQUE is in the middle of MATCH and CREATE it will match what it can, and create what is
missing. CREATE UNIQUE will always make the least change possible to the graph if it can use parts
the existing graph, it will.
of
Another difference to MATCH is that CREATE UNIQUE assumes the pattern to be unique. If multiple
matching subgraphs are found an error will be generated.
Tip
In the CREATE UNIQUE clause, patterns are used a lot. Read Section8.5, Patterns for an
introduction.
The examples start out with the following data set:
nam e = 'root '
nam e = 'A'
nam e = 'B'
KNOWS
nam e = 'C'
11.2.1.Create unique nodes

Create node if missing
If the pattern described needs a node, and it cant be matched, a new node will be created.
Query.
MATCH (root { name: 'root' })
CREATE UNIQUE (root)-[:LOVES]-(someone)
RETURN someone
The root node doesnt have any LOVES relationships, and so a node is created, and also a relationship to
that node.
Result
someone
Node[5]{}
1 row
Nodes created: 1
162
Writing Clauses
Create nodes with values
The pattern described can also contain values on the node. These are given using the following syntax:
prop : <expression>.
Query.
CREATE UNIQUE (root)-[:X]-(leaf { name:'D' })
RETURN leaf
No node connected with the root node has the name D, and so a new node is created to match the
pattern.
Result
leaf
Node[5]{name:"D"}
1 row
Nodes created: 1
Properties set: 1
Create labeled node if missing
If the pattern described needs a labeled node and there is none with the given labels, Cypher will
create a new one.
Query.
MATCH (a { name: 'A' })
CREATE UNIQUE (a)-[:KNOWS]-(c:blue)
RETURN c
The A node is connected in a KNOWS relationship to the c node, but since C doesnt have the :blue label,
a new node labeled as :blue is created along with a KNOWS relationship from A to it.
Result
c
Node[5]{}
1 row
Nodes created: 1
Labels added: 1
11.2.2.Create unique relationships

Create relationship if it is missing
CREATE UNIQUE is used to describe the pattern that should be found or created.
Query.
MATCH (lft { name: 'A' }),(rgt)
WHERE rgt.name IN ['B', 'C']
CREATE UNIQUE (lft)-[r:KNOWS]->(rgt)
163
Writing Clauses
RETURN r
The left node is matched agains the two right nodes. One relationship already exists and can be
matched, and the other relationship is created before it is returned.
Result
r
:KNOWS[5]{}
:KNOWS[3]{}
2 rows
Create relationship with values
Relationships to be created can also be matched on values.
Query.
CREATE UNIQUE (root)-[r:X { since:'forever' }]-()
RETURN r
In this example, we want the relationship to have a value, and since no such relationship can be found,
a new node and relationship are created. Note that since we are not interested in the created node, we
dont name it.
Result
r
:X[5]{since:"forever"}
1 row
Nodes created: 1
Properties set: 1
11.2.3.Describe complex pattern

The pattern described by CREATE UNIQUE can be separated by commas, just like in MATCH and CREATE.
Query.
CREATE UNIQUE (root)-[:FOO]->(x),(root)-[:BAR]->(x)
RETURN x
This example pattern uses two paths, separated by a comma.

Result
x
Node[5]{}
1 row
Nodes created: 1
164
Writing Clauses
11.3.Merge
11.3.1.Introduction
Provisional Feature
is a new addition and a provisional feature of Cypher. Its scope and syntax are
subject to change. Therefore, please be aware that you may need to change queries that use
MERGE in the future.
MERGE
ensures that a pattern exists in the graph. Either the pattern already exists, or it needs to be
created.
MERGE
either matches existing nodes and binds them, or it creates new data and binds that. Its like a
combination of MATCH and CREATE that additionally allows you to specify what happens if the data was
matched or created.
MERGE
For example, you can specify that the graph must contain a node for a user with a certain name. If
there isnt a node with the correct name, a new node will be created and its name property set.
When using MERGE on full patterns, the behavior is that either the whole pattern matches, or the whole
pattern is created. MERGE will not partially use existing patterns its all or nothing. If partial matches
are needed, this can be accomplished by splitting a pattern up into multiple MERGE clauses.
As with MATCH, MERGE can match multiple occurrences of a pattern. If there are multiple matches, they
will all be passed on to later stages of the query.
The last part of MERGE is the ON CREATE and ON MATCH. These allow a query to express additional changes
to the properties of a node or relationship, depending on if the element was MATCHed in the database or
if it was CREATEd.
Figure11.1.Graph
Person
FATHER
Person
Person
Person
Person
ACTED_IN
DIRECTED
ACTED_IN
ACTED_IN
Movie
ACTED_IN
ACTED_IN
Movie


11.3.2.Merge nodes
Merge single node with a label
Merging a single node with a given label.
Query.
165
DIRECTED
Writing Clauses
MERGE (robert:Critic)
RETURN robert, labels(robert)
Because there are no nodes labeled Critic in the database, a new node is created.
Result
robert
labels(robert)
Node[8]{}
["Critic"]
1 row
Nodes created: 1
Labels added: 1
Merge single node with properties
Merging a single node with properties where not all properties match any existing node.
Query.
MERGE (charlie { name:'Charlie Sheen', age:10 })
RETURN charlie
A new node with the name Charlie Sheen will be created since not all properties matched the existing
Charlie Sheen node.
Result
charlie
Node[8]{name:"Charlie Sheen", age:10}
1 row
Nodes created: 1
Properties set: 2
Merge single node specifying both label and property
Merging a single node with both label and property matching an existing node.
Query.
MERGE (michael:Person { name:'Michael Douglas' })
RETURN michael
Michael Douglas will be matched and returned.

Result
michael
1 row
11.3.3.Use ON CREATE and ON MATCH

Merge with ON CREATE
Merge a node and set properties if the node needs to be created.
166
Writing Clauses
Query.
MERGE (keanu:Person { name:'Keanu Reeves' })
ON CREATE SET keanu.created = timestamp()
RETURN keanu
Creates the Keanu node, and sets a timestamp on creation time.

Result
keanu
Node[8]{name:"Keanu Reeves", created:1386514444417}
1 row
Nodes created: 1
Properties set: 2
Labels added: 1
Merge with ON MATCH
Merging nodes and setting properties on found nodes.
Query.
MERGE (person:Person)
ON MATCH SET person.found = TRUE RETURN person
Finds all the Person nodes, sets a property on them, and returns them.
Result
person
Node[0]{name:"Oliver Stone", found:true}
Node[1]{name:"Charlie Sheen", found:true}
Node[2]{name:"Martin Sheen", found:true}
Node[5]{name:"Rob Reiner", found:true}
Node[6]{name:"Michael Douglas", found:true}
5 rows
Properties set: 5
Merge with ON CREATE and ON MATCH
Merge a node and set properties if the node needs to be created.
Query.
MERGE (keanu:Person { name:'Keanu Reeves' })
ON CREATE SET keanu.created = timestamp()
ON MATCH SET keanu.lastSeen = timestamp()
RETURN keanu
The query creates the Keanu node, and sets a timestamp on creation time. If Keanu already existed, a
different property would have been set.
167
Writing Clauses
Result
keanu
Node[8]{created:1386514447501, name:"Keanu Reeves"}
1 row
Nodes created: 1
Properties set: 2
Labels added: 1
11.3.4.Merge relationships
Merge on a relationship
MERGE
can be used to match or create a relationship.
Query.
MATCH (charlie:Person { name:'Charlie Sheen' }),(wallStreet:Movie { title:'Wall Street' })
MERGE (charlie)-[r:ACTED_IN]->(wallStreet)
RETURN r
Charlie Sheen had already been marked as acting on Wall Street, so the existing relationship is found
and returned. Note that in order to match or create a relationship when using MERGE, at least one bound
node must be specified, which is done via the MATCH clause in the above example.
Result
r
:ACTED_IN[0]{}
1 row
Merge on multiple relationships
When MERGE is used on a whole pattern, either everything matches, or everything is created.
Query.
MATCH (oliver:Person { name:'Oliver Stone' }),(reiner:Person { name:'Rob Reiner' })
MERGE (oliver)-[:DIRECTED]->(movie:Movie)<-[:ACTED_IN]-(reiner)
RETURN movie
In our example graph, Oliver Stone and Rob Reiner have never worked together. When we try to MERGE
a movie between them, Cypher will not use any of the existing movies already connected to either
person. Instead, a new movie node is created.
Result
movie
Node[8]{}
1 row
Nodes created: 1
Labels added: 1
168
Writing Clauses
11.3.5.Using unique constraints with MERGE

Cypher prevents getting conflicting results from MERGE when using patterns that involve uniqueness
constrains. In this case, there must be at most one node that matches that pattern.
For example, given two uniqueness constraints on :Person(id) and :Person(ssn): then a query such as
MERGE (n:Person {id: 12, ssn: 437}) will fail, if there are two different nodes (one with id 12 and one
with ssn 437) or if there is only one node with only one of the properties. In other words, there must
be exactly one node that matches the pattern, or no matching nodes.
Note that the following examples assume the existence of uniqueness constraints that have been
created using:
CREATE CONSTRAINT ON (n:Person) ASSERT n.name IS UNIQUE;
CREATE CONSTRAINT ON (n:Person) ASSERT n.role IS UNIQUE;
Merge using unique constraints creates a new node if no node is found

Merge using unique constraints creates a new node if no node is found.
Query.
MERGE (laurence:Person { name: 'Laurence Fishburne' })
RETURN laurence
The query creates the laurence node. If laurence already existed, merge would just return the existing
node.
Result
laurence
Node[8]{name:"Laurence Fishburne"}
1 row
Nodes created: 1
Properties set: 1
Labels added: 1
Merge using unique constraints matches an existing node
Merge using unique constraints matches an existing node.
Query.
MERGE (oliver:Person { name:'Oliver Stone' })
RETURN oliver
The oliver node already exists, so merge just returns it.

Result
oliver
1 row
Merge with unique constraints and partial matches
Merge using unique constraints fails when finding partial matches.
Query.
169
Writing Clauses
MERGE (michael:Person { name:'Michael Douglas', role:'Gordon Gekko' })
RETURN michael
While there is a matching unique michael node with the name Michael Douglas, there is no unique
node with the role of Gordon Gekko and merge fails to match.
Error message.
Merge did not find a matching node and can not create a new node due to conflicts
with both existing and missing unique nodes. The conflicting constraints are on:
:Person.name and :Person.role
Merge with unique constraints and conflicting matches

Merge using unique constraints fails when finding conflicting matches.
Query.
MERGE (oliver:Person { name:'Oliver Stone', role:'Gordon Gekko' })
RETURN oliver
While there is a matching unique oliver node with the name Oliver Stone, there is also another unique
node with the role of Gordon Gekko and merge fails to match.
Error message.
Merge did not find a matching node and can not create a new node due to conflicts
with both existing and missing unique nodes. The conflicting constraints are on:
:Person.name and :Person.role
11.3.6.Using map parameters with MERGE

MERGE does not support map parameters like for example CREATE does. To use map parameters with
MERGE, it is necessary to explicitly use the expected properties, like in the following example. For more
information on parameters, see Section7.5, Parameters.

Parameters.
{
"param" : {
"name" : "Keanu Reeves",
"role" : "Neo"
}
}
Query.
MERGE (oliver:Person { name: { param }.name, role: { param }.role })
RETURN oliver
Result
oliver
Node[8]{name:"Keanu Reeves", role:"Neo"}
1 row
Nodes created: 1
Properties set: 2
Labels added: 1
170
Writing Clauses
11.4.Set
Updating labels on nodes and properties on nodes and relationships is done with the SET clause. SET
can also be used with maps from parameters to set properties.
Note
Setting labels on a node is an idempotent operations if you try to set a label on a node
that already has that label on it, nothing happens. The query statistics will tell you if
something needed to be done or not.
The examples use this graph as a starting point:
nam e = 'St efan'
KNOWS
Swedish
nam e = 'Andres'
age = 36
awesom e = t rue
nam e = 'Em il'
KNOWS
KNOWS
nam e = 'Pet er'

age = 34
11.4.1.Set a property
To set a property on a node or relationship, use SET.
Query.
MATCH (n { name: 'Andres' })
SET n.surname = 'Taylor'
RETURN n
The newly changed node is returned by the query.

Result
n
1 row
Properties set: 1
11.4.2.Remove a property
Normally you remove a property by using REMOVE, but its sometimes handy to do it using the SET
command. One example is if the property comes from a parameter.
Query.
171
Writing Clauses
SET n.name = NULL RETURN n
The node is returned by the query, and the name property is now missing.
Result
n
Node[3]{age:36, awesome:true}
1 row
Properties set: 1
11.4.3.Copying properties between nodes and relationships

You can also use SET to copy all properties from one graph element to another. Remember that doing
this will remove all other properties on the receiving graph element.
Query.
MATCH (at { name: 'Andres' }),(pn { name: 'Peter' })
SET at = pn
RETURN at, pn
The Andres node has had all its properties replaced by the properties in the Peter node.
Result
at
pn
Node[3]{age:34, name:"Peter"}
1 row
Properties set: 3
11.4.4.Set a property using a parameter

Use a parameter to give the value of a property.
Parameters.
{
"surname" : "Taylor"
}
Query.
SET n.surname = { surname }
RETURN n
The Andres node has got an surname added.

Result
n
1 row
Properties set: 1
172
Writing Clauses
11.4.5.Set all properties using a parameter

This will replace all existing properties on the node with the new set provided by the parameter.
Parameters.
{
"props" : {
"name" : "Andres",
}
}
Query.
SET n = { props }
RETURN n
The Andres node has had all its properties replaced by the properties in the props parameter.
Result
n
Node[3]{position:"Developer", name:"Andres"}
1 row
Properties set: 4
11.4.6.Set a label on a node

To set a label on a node, use SET.
Query.
MATCH (n { name: 'Stefan' })
SET n :German
RETURN n
The newly labeled node is returned by the query.

Result
n
Node[1]{name:"Stefan"}
1 row
Labels added: 1
11.4.7.Set multiple labels on a node

To set multiple labels on a node, use SET and separate the different labels using :.
Query.
MATCH (n { name: 'Emil' })
SET n :Swedish:Bossman
RETURN n
The newly labeled node is returned by the query.

173
Writing Clauses
Result
n
Node[0]{name:"Emil"}
1 row
Labels added: 2
174
Writing Clauses
11.5.Delete
Deleting graph elements nodes and relationships, is done with DELETE.
For removing properties and labels, see Section11.6, Remove.
The examples start out with the following database:
nam e = 'Andres'
age = 36
KNOWS
nam e = 'Tobias'
age = 25
KNOWS
nam e = 'Pet er'

age = 34
11.5.1.Delete single node

To delete a node, use the DELETE clause.
Query.
MATCH (n { name: 'Peter' })
DELETE n
Result
(empty result)
Nodes deleted: 1
11.5.2.Delete a node and connected relationships

If you are trying to delete a node with relationships on it, you have to delete these as well.
Query.
MATCH (n { name: 'Andres' })-[r]-()
DELETE n, r
Result
(empty result)
Nodes deleted: 1
11.5.3.Delete all nodes and relationships

This query isnt for deleting large amounts of data, but is nice when playing around with small
example data sets.
Query.
MATCH (n)
175
Writing Clauses
OPTIONAL MATCH (n)-[r]-()
DELETE n,r
Result
(empty result)
Nodes deleted: 3
176
Writing Clauses
11.6.Remove
Removing properties and labels from graph elements is done using REMOVE.
For deleting nodes and relationships, see Section11.5, Delete.
Note
Removing labels from a node is an idempotent operation: If you try to remove a label from
a node that does not have that label on it, nothing happens. The query statistics will tell you
if something needed to be done or not.
The examples start out with the following database:
Swedish
nam e = 'Andres'
age = 36
KNOWS
Swedish
KNOWS
Swedish, Germ an
nam e = 'Tobias'
age = 25
nam e = 'Pet er'

age = 34
11.6.1.Remove a property
Neo4j doesnt allow storing null in properties. Instead, if no value exists, the property is just not there.
So, to remove a property value on a node or a relationship, is also done with REMOVE.
Query.
MATCH (andres { name: 'Andres' })
REMOVE andres.age
RETURN andres
The node is returned, and no property age exists on it.

Result
andres
Node[2]{name:"Andres"}
1 row
Properties set: 1
11.6.2.Remove a label from a node

To remove labels, you use REMOVE.
Query.
REMOVE n:German
RETURN n
177
Writing Clauses
Result
n
1 row
Labels removed: 1
11.6.3.Removing multiple labels

To remove multiple labels, you use REMOVE.
Query.
REMOVE n:German:Swedish
RETURN n
Result
n
1 row
Labels removed: 2
178
Writing Clauses
11.7.Foreach
Collections and paths are key concepts in Cypher. To use them for updating data, you can use the
FOREACH construct. It allows you to do updating commands on elements in a collection a path, or a
collection created by aggregation.
The identifier context inside of the foreach parenthesis is separate from the one outside it. This means
that if you CREATE a node identifier inside of a FOREACH, you will not be able to use it outside of the
foreach statement, unless you match to find it.
Inside of the FOREACH parentheses, you can do any of the updating commands CREATE, CREATE UNIQUE,
DELETE, and FOREACH.
Figure11.2.Data for the examples
nam e = 'A'
KNOWS
nam e = 'B'
KNOWS
nam e = 'C'
KNOWS
nam e = 'D'
11.7.1.Mark all nodes along a path

This query will set the property marked to true on all nodes along a path.
Query.
MATCH p =(begin)-[*]->(END )
WHERE begin.name='A' AND END .name='D'
FOREACH (n IN nodes(p)| SET n.marked = TRUE )
Nothing is returned from this query, but four properties are set.
Result
(empty result)
Properties set: 4
179
Chapter12.Functions
This chapter contains information on all functions in Cypher. Note that related information exists in
Section8.1, Operators.
Note
Most functions in Cypher will return NULL if an input parameter is NULL.
180
Functions
12.1.Predicates
Predicates are boolean functions that return true or false for a given set of input. They are most
commonly used to filter out subgraphs in the WHERE part of a query.
See also Section8.1.2, Comparison operators.
Figure12.1.Graph
foo, bar
nam e = 'Alice'
age = 38
eyes = 'brown'
KNOWS
nam e = 'Bob'
age = 25
eyes = 'blue'
MARRIED
KNOWS
nam e = 'Charlie'
age = 53
eyes = 'green'
KNOWS
KNOWS
Spouse
nam e = 'Eskil'
age = 41
eyes = 'blue'
array = [ 'one', 't wo', 't hree']
nam e = 'Daniel'
age = 54
eyes = 'brown'
12.1.1.ALL
Tests whether a predicate holds for all element of this collection collection.
Syntax: ALL(identifier in collection WHERE predicate)
Arguments:
collection: An expression that returns a collection
identifier: This is the identifier that can be used from the predicate.
predicate: A predicate that is tested against all items in the collection.
Query.
MATCH p=(a)-[*1..3]->(b)
WHERE a.name='Alice' AND b.name='Daniel' AND ALL (x IN nodes(p) WHERE x.age > 30)
RETURN p
All nodes in the returned paths will have an age property of at least 30.
Result
p
[Node[2]{name:"Alice", age:38, eyes:"brown"}, :KNOWS[1]{}, Node[4]{name:"Charlie", age:53,
eyes:"green"}, :KNOWS[3]{}, Node[0]{name:"Daniel", age:54, eyes:"brown"}]
1 row
181
Functions
12.1.2.ANY
Tests whether a predicate holds for at least one element in the collection.
Syntax: ANY(identifier in collection WHERE predicate)
Arguments:
Query.
MATCH (a)
WHERE a.name='Eskil' AND ANY (x IN a.array WHERE x = "one")
RETURN a
All nodes in the returned paths has at least one one value set in the array property named array.
Result
a
Node[1]{name:"Eskil", age:41, eyes:"blue", array:["one", "two", "three"]}
1 row
12.1.3.NONE
Returns true if the predicate holds for no element in the collection.
Syntax: NONE(identifier in collection WHERE predicate)
Arguments:
Query.
MATCH p=(n)-[*1..3]->(b)
WHERE n.name='Alice' AND NONE (x IN nodes(p) WHERE x.age = 25)
RETURN p
No nodes in the returned paths has a age property set to 25.

Result
p
eyes:"green"}]
eyes:"green"}, :KNOWS[3]{}, Node[0]{name:"Daniel", age:54, eyes:"brown"}]
2 rows
12.1.4.SINGLE
Returns true if the predicate holds for exactly one of the elements in the collection.
182
Functions
Syntax: SINGLE(identifier in collection WHERE predicate)
Arguments:
Query.
MATCH p=(n)-->(b)
WHERE n.name='Alice' AND SINGLE (var IN nodes(p) WHERE var.eyes = "blue")
RETURN p
Exactly one node in every returned path will have the eyes property set to "blue".
Result
p
[Node[2]{name:"Alice", age:38, eyes:"brown"}, :KNOWS[0]{}, Node[3]{name:"Bob", age:25, eyes:"blue"}]
1 row
183
Functions
12.2.Scalar functions
Scalar functions return a single value.
Figure12.2.Graph
foo, bar
nam e = 'Alice'
age = 38
eyes = 'brown'
KNOWS
nam e = 'Bob'
age = 25
eyes = 'blue'
MARRIED
KNOWS
nam e = 'Charlie'
age = 53
eyes = 'green'
KNOWS
KNOWS
Spouse
nam e = 'Eskil'
age = 41
eyes = 'blue'
nam e = 'Daniel'
age = 54
eyes = 'brown'
12.2.1.LENGTH
To return or filter on the length of a collection, use the LENGTH() function.
Syntax: LENGTH( collection )
Arguments:
Query.
MATCH p=(a)-->(b)-->(c)
WHERE a.name='Alice'
RETURN length(p)
The length of the path p is returned by the query.

Result
length(p)
2
2
2
3 rows
12.2.2.TYPE
Returns a string representation of the relationship type.
184
Functions
Syntax: TYPE( relationship )
Arguments:
relationship: A relationship.
Query.
MATCH (n)-[r]->()
WHERE n.name='Alice'
RETURN type(r)
The relationship type of r is returned by the query.

Result
type(r)
"KNOWS"
"KNOWS"
2 rows
12.2.3.ID
Returns the id of the relationship or node.
Syntax: ID( property-container )
Arguments:
property-container: A node or a relationship.
Query.
MATCH (a)
RETURN id(a)
This returns the node id for three nodes.

Result
id(a)
0
1
2
3
4
5 rows
12.2.4.COALESCE
Returns the first non-NULL value in the list of expressions passed to it. In case all arguments are NULL,
NULL will be returned.
Syntax: COALESCE( expression [, expression]* )
185
Functions
Arguments:
expression: The expression that might return NULL.
Query.
MATCH (a)
RETURN coalesce(a.hairColor, a.eyes)
Result
coalesce(a.hairColor, a.eyes)
"brown"
1 row
12.2.5.HEAD
HEAD
returns the first element in a collection.
Syntax: HEAD( expression )

Arguments:
expression: This expression should return a collection of some kind.
Query.
MATCH (a)
WHERE a.name='Eskil'
RETURN a.array, head(a.array)
The first node in the path is returned.

Result
a.array
head(a.array)
["one", "two", "three"]
"one"
1 row
12.2.6.LAST
LAST
returns the last element in a collection.
Syntax: LAST( expression )

Arguments:
Query.
MATCH (a)
RETURN a.array, last(a.array)
The last node in the path is returned.

186
Functions
Result
a.array
last(a.array)
"three"
1 row
12.2.7.TIMESTAMP
returns the difference, measured in milliseconds, between the current time and midnight,
January 1, 1970 UTC. It will return the same value during the whole one query, even if the query is a
long running one.
TIMESTAMP
Syntax: TIMESTAMP()
Arguments:
Query.
RETURN timestamp()
The time in milliseconds is returned.

Result
timestamp()
1386514453776
1 row
12.2.8.STARTNODE
STARTNODE
returns the starting node of a relationship
Syntax: STARTNODE( relationship )

Arguments:
relationship: An expression that returns a relationship
Query.
MATCH (x:foo)-[r]-()
RETURN startNode(r)
Result
startNode(r)
Node[2]{name:"Alice", age:38, eyes:"brown"}
Node[2]{name:"Alice", age:38, eyes:"brown"}
2 rows
12.2.9.ENDNODE
ENDNODE
returns the end node of a relationship
Syntax: ENDNODE( relationship )

Arguments:
187
Functions
relationship: An expression that returns a relationship
Query.
MATCH (x:foo)-[r]-()
RETURN endNode(r)
Result
endNode(r)
Node[3]{name:"Bob", age:25, eyes:"blue"}
Node[4]{name:"Charlie", age:53, eyes:"green"}
2 rows
188
Functions
12.3.Collection functions
Collection functions return collections of things nodes in a path, and so on.
See also Section8.1.5, Collection operators.
Figure12.3.Graph
foo, bar
nam e = 'Alice'
age = 38
eyes = 'brown'
KNOWS
nam e = 'Bob'
age = 25
eyes = 'blue'
MARRIED
KNOWS
nam e = 'Charlie'
age = 53
eyes = 'green'
KNOWS
KNOWS
Spouse
nam e = 'Eskil'
age = 41
eyes = 'blue'
nam e = 'Daniel'
age = 54
eyes = 'brown'
12.3.1.NODES
Returns all nodes in a path.
Syntax: NODES( path )
Arguments:
path: A path.
Query.
MATCH p=(a)-->(b)-->(c)
WHERE a.name='Alice' AND c.name='Eskil'
RETURN nodes(p)
All the nodes in the path p are returned by the example query.
Result
nodes(p)
[Node[2]{name:"Alice", age:38, eyes:"brown"}, Node[3]{name:"Bob", age:25, eyes:"blue"}, Node[1]
{name:"Eskil", age:41, eyes:"blue", array:["one", "two", "three"]}]
1 row
12.3.2.RELATIONSHIPS
Returns all relationships in a path.
189
Functions
Syntax: RELATIONSHIPS( path )
Arguments:
path: A path.
Query.
MATCH p=(a)-->(b)-->(c)
WHERE a.name='Alice' AND c.name='Eskil'
RETURN relationships(p)
All the relationships in the path p are returned.

Result
relationships(p)
[:KNOWS[0]{}, :MARRIED[4]{}]
1 row
12.3.3.LABELS
Returns a collection of string representations for the labels attached to a node.
Syntax: LABELS( node )
Arguments:
node: Any expression that returns a single node
Query.
MATCH (a)
RETURN labels(a)
The labels of n is returned by the query.

Result
labels(a)
["foo", "bar"]
1 row
12.3.4.EXTRACT
To return a single property, or the value of a function from a collection of nodes or relationships, you
can use EXTRACT. It will go through a collection, run an expression on every element, and return the
results in an collection with these values. It works like the map method in functional languages such as
Lisp and Scala.
Syntax: EXTRACT( identifier in collection | expression )
Arguments:
identifier: The closure will have an identifier introduced in its context. Here you decide which
identifier to use.
190
Functions
expression: This expression will run once per value in the collection, and produces the result
collection.
Query.
MATCH p=(a)-->(b)-->(c)
WHERE a.name='Alice' AND b.name='Bob' AND c.name='Daniel'
RETURN extract(n IN nodes(p)| n.age) AS extracted
The age property of all nodes in the path are returned.

Result
extracted
[38, 25, 54]
1 row
12.3.5.FILTER
FILTER
returns all the elements in a collection that comply to a predicate.
Syntax: FILTER(identifier in collection WHERE predicate)

Arguments:
Query.
MATCH (a)
RETURN a.array, filter(x IN a.array WHERE length(x)= 3)
This returns the property named array and a list of values in it, which have the length 3.
Result
a.array
filter(x in a.array WHERE length(x) = 3)
["one", "two"]
1 row
12.3.6.TAIL
TAIL
returns all but the first element in a collection.
Syntax: TAIL( expression )

Arguments:
Query.
MATCH (a)
RETURN a.array, tail(a.array)
191
Functions
This returns the property named array and all elements of that property except the first one.
Result
a.array
tail(a.array)
["two", "three"]
1 row
12.3.7.RANGE
Returns numerical values in a range with a non-zero step value step. Range is inclusive in both ends.
Syntax: RANGE( start, end [, step] )
Arguments:
start: A numerical expression.
end: A numerical expression.
step: A numerical expression.
Query.
RETURN range(0,10), range(2,18,3)
Two lists of numbers are returned.

Result
range(0,10)
range(2,18,3)
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
[2, 5, 8, 11, 14, 17]
1 row
12.3.8.REDUCE
To run an expression against individual elements of a collection, and store the result of the expression
in an accumulator, you can use REDUCE. It will go through a collection, run an expression on every
element, storing the partial result in the accumulator. It works like the fold or reduce method in
functional languages such as Lisp and Scala.
Syntax: REDUCE( accumulator = initial, identifier in collection | expression )
Arguments:
accumulator: An identifier that will hold the result and the partial results as the collection is iterated
initial: An expression that runs once to give a starting value to the accumulator
identifier: The closure will have an identifier introduced in its context. Here you decide which
identifier to use.
expression: This expression will run once per value in the collection, and produces the result value.
Query.
MATCH p=(a)-->(b)-->(c)
WHERE a.name='Alice' AND b.name='Bob' AND c.name='Daniel'
RETURN reduce(totalAge = 0, n IN nodes(p)| totalAge + n.age) AS reduction
192
Functions
The age property of all nodes in the path are summed and returned as a single value.
Result
reduction
117
1 row
193
Functions
12.4.Mathematical functions
These functions all operate on numerical expressions only, and will return an error if used on any
other values.
See also Section8.1.1, Mathematical operators.
Figure12.4.Graph
foo, bar
nam e = 'Alice'
age = 38
eyes = 'brown'
KNOWS
nam e = 'Bob'
age = 25
eyes = 'blue'
MARRIED
KNOWS
nam e = 'Charlie'
age = 53
eyes = 'green'
KNOWS
KNOWS
Spouse
nam e = 'Eskil'
age = 41
eyes = 'blue'
nam e = 'Daniel'
age = 54
eyes = 'brown'
12.4.1.ABS
ABS
returns the absolute value of a number.
Syntax: ABS( expression )

Arguments:
expression: A numeric expression.
Query.
MATCH (a),(e)
WHERE a.name = 'Alice' AND e.name = 'Eskil'
RETURN a.age, e.age, abs(a.age - e.age)
The absolute value of the age difference is returned.

Result
a.age
e.age
abs(a.age - e.age)
38
41
3. 0
1 row
12.4.2.ACOS
ACOS
returns the arccosine of the expression, in radians.
Syntax: ACOS( expression )

194
Functions
Arguments:
Query.
RETURN acos(0.5)
The arccosine of 0.5.

Result
acos(0.5)
1. 0471975511965979
1 row
12.4.3.ASIN
ASIN
returns the arcsine of the expression, in radians.
Syntax: ASIN( expression )

Arguments:
Query.
RETURN asin(0.5)
The arcsine of 0.5.

Result
asin(0.5)
0. 5235987755982989
1 row
12.4.4.ATAN
ATAN
returns the arctangent of the expression, in radians.
Syntax: ATAN( expression )

Arguments:
Query.
RETURN atan(0.5)
The arctangent of 0.5.

Result
atan(0.5)
0. 4636476090008061
1 row
195
Functions
12.4.5.ATAN2
ATAN2
returns the arctangent2 of a set of coordinates, in radians.
Syntax: ATAN2( expression , expression)

Arguments:
expression: A numeric expression for y.
expression: A numeric expression for x.
Query.
RETURN atan2(0.5, 0.6)
The arctangent2 of 0.5, 0.6.

Result
atan2(0.5, 0.6)
0. 6947382761967033
1 row
12.4.6.COS
COS
returns the cosine of the expression.
Syntax: COS( expression )

Arguments:
Query.
RETURN cos(0.5)
The cosine of 0.5 is returned.

Result
cos(0.5)
0. 8775825618903728
1 row
12.4.7.COT
COT
returns the cotangent of the expression.
Syntax: COT( expression )

Arguments:
Query.
RETURN cot(0.5)
The cotangent of 0.5 is returned.

196
Functions
Result
cot(0.5)
1. 830487721712452
1 row
12.4.8.DEGREES
DEGREES
converts radians to degrees.
Syntax: DEGREES( expression )

Arguments:
Query.
RETURN degrees(3.14159)
The number of degrees in something close to pi.

Result
degrees(3.14159)
179. 99984796050427
1 row
12.4.9.E
E
returns the constant, e.
Syntax: E( expression )
Arguments:
Query.
RETURN e()
The constant e is returned (the base of natural log).

Result
e()
2. 718281828459045
1 row
12.4.10.EXP
EXP
returns the value e raised to the power of the expression.
Syntax: EXP( expression )

Arguments:
Query.
197
Functions
RETURN exp(2)
The exp of 2 is returned: e2.

Result
exp(2)
7. 38905609893065
1 row
12.4.11.FLOOR
FLOOR
returns the greatest integer less than or equal to the expression.
Syntax: FLOOR( expression )

Arguments:
Query.
RETURN floor(0.9)
The floor of 0.9 is returned.

Result
floor(0.9)
0. 0
1 row
12.4.12.HAVERSIN
HAVERSIN
returns the half versine of the expression.
Syntax: HAVERSIN( expression )

Arguments:
Query.
RETURN haversin(0.5)
The haversine of 0.5 is returned.

Result
haversin(0.5)
0. 06120871905481362
1 row
Spherical distance using the haversin function
The haversin function may be used to compute the distance on the surface of a sphere between two
points (each given by their latitude and longitude). In this example the spherical distance (in km)
198
Functions
between Berlin in Germany (at lat 52.5, lon 13.4) and San Mateo in California (at lat 37.5, lon -122.3)
is calculated using an average earth radius of 6371 km.
Query.
CREATE (ber:City { lat: 52.5, lon: 13.4 }),(sm:City { lat: 37.5, lon: -122.3 })
RETURN 2 * 6371 * asin(sqrt(haversin(radians(sm.lat - ber.lat))+ cos(radians(sm.lat))*
cos(radians(ber.lat))* haversin(radians(sm.lon - ber.lon)))) AS dist
The distance between Berlin and San Mateo is returned (about 9129 km).
Result
dist
9129. 969740051658
1 row
Nodes created: 2
Properties set: 4
Labels added: 2
12.4.13.LOG
LOG
returns the natural logarithm of the expression.
Syntax: LOG( expression )

Arguments:
Query.
RETURN log(27)
The log of 27 is returned.

Result
log(27)
3. 295836866004329
1 row
12.4.14.LOG10
LOG10
returns the base 10 logarithm of the expression.
Syntax: LOG10( expression )

Arguments:
Query.
RETURN log10(27)
The log10 of 27 is returned.

199
Functions
Result
log10(27)
1. 4313637641589874
1 row
12.4.15.PI
PI
returns the mathmatical constant pi.
Syntax: PI()
Arguments:
Query.
RETURN pi()
The constant pi is returned.

Result
pi()
3. 141592653589793
1 row
12.4.16.RADIANS
RADIANS
converts degrees to radians.
Syntax: RADIANS( expression )

Arguments:
Query.
RETURN radians(180)
The number of radians in 180 is returned (pi).

Result
radians(180)
3. 141592653589793
1 row
12.4.17.RAND
RAND
returns a random double between 0 and 1.0.
Syntax: RAND( expression )

Arguments:
Query.
200
Functions
RETURN rand() AS x1
A random number is returned.

Result
x1
0. 16105338985811957
1 row
12.4.18.ROUND
ROUND
returns the numerical expression, rounded to the nearest integer.
Syntax: ROUND( expression )

Arguments:
expression: A numerical expression.
Query.
RETURN round(3.141592)
Result
round(3.141592)
3
1 row
12.4.19.SIGN
SIGN returns the signum of a
1 for any positive number.
number zero if the expression is zero, -1 for any negative number, and
Syntax: SIGN( expression )

Arguments:
expression: A numerical expression
Query.
RETURN sign(-17), sign(0.1)
Result
sign(-17)
sign(0.1)
-1. 0
1. 0
1 row
12.4.20.SIN
SIN
returns the sine of the expression.
Syntax: SIN( expression )

Arguments:
201
Functions
Query.
RETURN sin(0.5)
The sine of 0.5 is returned.

Result
sin(0.5)
0. 479425538604203
1 row
12.4.21.SQRT
SQRT
returns the square root of a number.
Syntax: SQRT( expression )

Arguments:
expression: A numerical expression
Query.
RETURN sqrt(256)
Result
sqrt(256)
16. 0
1 row
12.4.22.TAN
TAN
returns the tangent of the expression.
Syntax: TAN( expression )

Arguments:
Query.
RETURN tan(0.5)
The tangent of 0.5 is returned.

Result
tan(0.5)
0. 5463024898437905
1 row
202
Functions
12.5.String functions
These functions all operate on string expressions only, and will return an error if used on any other
values. Except STR(), which converts to strings.
See also Section8.1.4, String operators.
Figure12.5.Graph
foo, bar
nam e = 'Alice'
age = 38
eyes = 'brown'
KNOWS
nam e = 'Bob'
age = 25
eyes = 'blue'
MARRIED
KNOWS
nam e = 'Charlie'
age = 53
eyes = 'green'
KNOWS
KNOWS
Spouse
nam e = 'Eskil'
age = 41
eyes = 'blue'
nam e = 'Daniel'
age = 54
eyes = 'brown'
12.5.1.STR
returns a string representation of the expression.
STR
Syntax: STR( expression )

Arguments:
expression: An expression that returns anything
Query.
RETURN str(1)
Result
str(1)
"1"
1 row
12.5.2.REPLACE
returns a string with the search string replaced by the replace string. It replaces all
occurrences.
REPLACE
Syntax: REPLACE( original, search, replace )

Arguments:
203
Functions
original: An expression that returns a string
search: An expression that returns a string to search for
replace: An expression that returns the string to replace the search string with
Query.
RETURN replace("hello", "l", "w")
Result
replace("hello", "l", "w")
"hewwo"
1 row
12.5.3.SUBSTRING
returns a substring of the original, with a 0-based index start and length. If length is omitted,
it returns a substring from start until the end of the string.
SUBSTRING
Syntax: SUBSTRING( original, start [, length] )

Arguments:
start: An expression that returns a positive number
length: An expression that returns a positive number
Query.
RETURN substring("hello", 1, 3), substring("hello", 2)
Result
substring("hello", 1, 3)
substring("hello", 2)
"ell"
"llo"
1 row
12.5.4.LEFT
LEFT
returns a string containing the left n characters of the original string.
Syntax: LEFT( original, length )

Arguments:
n: An expression that returns a positive number
Query.
RETURN left("hello", 3)
Result
left("hello", 3)
"hel"
1 row
204
Functions
12.5.5.RIGHT
returns a string containing the right n characters of the original string.
RIGHT
Syntax: RIGHT( original, length )

Arguments:
n: An expression that returns a positive number
Query.
RETURN right("hello", 3)
Result
right("hello", 3)
"llo"
1 row
12.5.6.LTRIM
LTRIM
returns the original string with whitespace removed from the left side.
Syntax: LTRIM( original )

Arguments:
Query.
RETURN ltrim("
hello")
Result
ltrim(" hello")
"hello"
1 row
12.5.7.RTRIM
RTRIM
returns the original string with whitespace removed from the right side.
Syntax: RTRIM( original )

Arguments:
Query.
RETURN rtrim("hello
")
Result
rtrim("hello ")
"hello"
1 row
205
Functions
12.5.8.TRIM
TRIM
returns the original string with whitespace removed from both sides.
Syntax: TRIM( original )

Arguments:
Query.
RETURN trim("
hello
")
Result
trim(" hello ")
"hello"
1 row
12.5.9.LOWER
LOWER
returns the original string in lowercase.
Syntax: LOWER( original )

Arguments:
Query.
RETURN lower("HELLO")
Result
lower("HELLO")
"hello"
1 row
12.5.10.UPPER
UPPER
returns the original string in uppercase.
Syntax: UPPER( original )

Arguments:
Query.
RETURN upper("hello")
Result
upper("hello")
"HELLO"
1 row
206
Chapter13.Schema
Neo4j 2.0 introduced an optional schema for the graph, based around the concept of labels. Labels are
used in the specification of indexes, and for defining constraints on the graph. Together, indexes and
constraints are the schema of the graph. Cypher includes data definition language (DDL) statements
for manipulating the schema.
207
Schema
13.1.Indexes
Cypher allows the creation of indexes over a property for all nodes that have a given label. These
indexes are automatically managed and kept up to date by the database whenever the graph is
changed.
13.1.1.Create index on a label

To create an index on a property for all nodes that have a label, use CREATE INDEX ON. Note that the
index is not immediately available, but will be created in the background. See Section3.7.1, Indexes
for details.
Query.
CREATE INDEX ON :Person(name)
Result
(empty result)
Indexes added: 1
13.1.2.Drop index on a label

To drop an index on all nodes that have a label, use the DROP INDEX clause.
Query.
DROP INDEX ON :Person(name)
Result
(empty result)
Indexes removed: 1
13.1.3.Use index
There is usually no need to specify which indexes to use in a query, Cypher will figure that out by
itself. For example the query below will use the Person(name) index, if it exists. If you for some reason
want to hint to specific indexes, see Section9.7, Using.
Query.
MATCH (n:Person { name: 'Andres' })
RETURN n
Result
n
(empty result)
0 row
208
Schema
13.2.Constraints
Neo4j helps enforce data integrity with the use of constraints.
You can use unique constraints to ensure that property values are unique for all nodes with a
specific label. Unique constraints do not mean that all nodes have to have a unique value for the
properties nodes without the property are not subject to this rule.
Remember that adding constraints is an atomic operation that can take a while all existing data has
to be scanned before Neo4j can turn the constraint on.
You can have multiple unique constraints for a given label.
Note that adding a uniqueness constraint on a property will also add an index on that property, so you
cannot add such an index separately. Cypher will use that index for lookups just like other indexes. If
you drop a constraint and still want an index on the property, you will have to create the index.
13.2.1.Create uniqueness constraint

To create a constraint that makes sure that your database will never contain more than one node with a
specific label and one property value, use the IS UNIQUE syntax.
Query.
CREATE CONSTRAINT ON (book:Book) ASSERT book.isbn IS UNIQUE
Result
(empty result)
Constraints added: 1
13.2.2.Drop uniqueness constraint

By using DROP CONSTRAINT, you remove a constraint from the database.
Query.
DROP CONSTRAINT ON (book:Book) ASSERT book.isbn IS UNIQUE
Result
(empty result)
Constraints removed: 1
13.2.3.Create a node that complies with constraints

Create a Book node with an isbn that isnt already in the database.
Query.
CREATE (book:Book { isbn: '1449356265', title: 'Graph Databases' })
Result
(empty result)
Nodes created: 1
Properties set: 2
Labels added: 1
209
Schema
13.2.4.Create a node that breaks a constraint

Create a Book node with an isbn that is already used in the database.
Query.
CREATE (book:Book { isbn: '1449356265', title: 'Graph Databases' })
Error message.
Node 0 already exists with label Book and property "isbn"=[1449356265]
210
Chapter14.From SQL to Cypher

This guide is for people who understand SQL. You can use that prior knowledge to quickly get going
with Cypher and start exploring Neo4j.
211
From SQL to Cypher
14.1.Start
SQL starts with the result you want we SELECT what we want and then declare how to source it. In
Cypher, the START clause is quite a different concept which specifies starting points in the graph from
which the query will execute.
From a SQL point of view, the identifiers in START are like table names that point to a set of nodes
or relationships. The set can be listed literally, come via parameters, or as I show in the following
example, be defined by an index look-up.
So in fact rather than being SELECT-like, the START clause is somewhere between the FROM and the WHERE
clause in SQL.
SQL Query.
SELECT *
FROM "Person"
WHERE name = 'Anakin'
NAME
ID
AGE
HAIR
Anakin
20
blonde
1 rows
Cypher Query.
START person=node:Person(name = 'Anakin')
RETURN person
person
Node[0]{name:"Anakin", id:1, age:20, hair:"blonde"}
1 row
Cypher allows multiple starting points. This should not be strange from a SQL perspective every
table in the FROM clause is another starting point.
212
From SQL to Cypher
14.2.Match
Unlike SQL which operates on sets, Cypher predominantly works on sub-graphs. The relational
equivalent is the current set of tuples being evaluated during a SELECT query.
The shape of the sub-graph is specified in the MATCH clause. The MATCH clause is analogous to the JOIN in
SQL. A normal ab relationship is an inner join between nodes a and b both sides have to have at
least one match, or nothing is returned.
Well start with a simple example, where we find all email addresses that are connected to the person
Anakin. This is an ordinary one-to-many relationship.
SQL Query.
SELECT "Email".*
FROM "Person"
JOIN "Email" ON "Person".id = "Email".person_id
WHERE "Person".name = 'Anakin'
ADDRESS
COMMENT
PERSON_ID
anakin@example. com
home
anakin@example. org
work
2 rows
Cypher Query.
START person=node:Person(name = 'Anakin')
MATCH person-[:email]->email
RETURN email
email
Node[6]{address:"anakin@example. com", comment:"home"}
Node[7]{address:"anakin@example. org", comment:"work"}
2 rows
There is no join table here, but if one is necessary the next example will show how to do that, writing
the pattern relationship like so: -[r:belongs_to]-> will introduce (the equivalent of) join table available
as the variable r. In reality this is a named relationship in Cypher, so were saying join Person to Group
via belongs_to. To illustrate this, consider this image, comparing the SQL model and Neo4j/Cypher.
213
From SQL to Cypher

And here are example queries:
SQL Query.
SELECT "Group".*, "Person_Group".*
FROM "Person"
JOIN "Person_Group" ON "Person".id = "Person_Group".person_id
JOIN "Group" ON "Person_Group".Group_id="Group".id
WHERE "Person".name = 'Bridget'
NAME
ID
BELONGS_TO_GROUP_ID
PERSON_ID
GROUP_ID
Admin
1 rows
Cypher Query.
START person=node:Person(name = 'Bridget')
MATCH person-[r:belongs_to]->group
RETURN group, r
group
Node[5]{name:"Admin", id:4}
:belongs_to[0]{}
1 row
An outer join <http://www.codinghorror.com/blog/2007/10/a-visual-explanation-of-sql-joins.html>
is just as easy. Add OPTIONAL before the match and its an optional relationship between nodes the
outer join of Cypher.
Whether its a left outer join, or a right outer join is defined by which side of the pattern has a starting
point. This example is a left outer join, because the bound node is on the left side:
SQL Query.
SELECT "Person".name, "Email".address
FROM "Person" LEFT
JOIN "Email" ON "Person".id = "Email".person_id
NAME
ADDRESS
Anakin
anakin@example. com
Anakin
anakin@example. org
Bridget
<null>
3 rows
Cypher Query.
START person=node:Person('name: *')
OPTIONAL MATCH person-[:email]->email
RETURN person.name, email.address
person.name
email.address
"Anakin"
"anakin@example. com"
3 rows
214
From SQL to Cypher

person.name
email.address
"Anakin"
"anakin@example. org"
"Bridget"
<null>
3 rows
Relationships in Neo4j are first class citizens its like the SQL tables are pre-joined with each other.
So, naturally, Cypher is designed to be able to handle highly connected data easily.
One such domain is tree structures anyone that has tried storing tree structures in SQL knows that
you have to work hard to get around the limitations of the relational model. There are even books on
the subject.
To find all the groups and sub-groups that Bridget belongs to, this query is enough in Cypher:
Cypher Query.
START person=node:Person('name: Bridget')
MATCH person-[:belongs_to*]->group
RETURN person.name, group.name
person.name
group.name
"Bridget"
"Admin"
"Bridget"
"Technichian"
"Bridget"
"User"
3 rows
The * after the relationship type means that there can be multiple hops across belongs_to relationships
between group and user. Some SQL dialects have recursive abilities, that allow the expression
of queries like this, but you may have a hard time wrapping your head around those. Expressing
something like this in SQL is hugely impractical if not practically impossible.
215
From SQL to Cypher
14.3.Where
This is the easiest thing to understand its the same animal in both languages. It filters out result
sets/subgraphs. Not all predicates have an equivalent in the other language, but the concept is the
same.
SQL Query.
SELECT *
FROM "Person"
WHERE "Person".age > 35 AND "Person".hair = 'blonde'
NAME
ID
AGE
HAIR
Bridget
40
blonde
1 rows
Cypher Query.
WHERE person.age > 35 AND person.hair = 'blonde'
RETURN person
person
Node[1]{name:"Bridget", id:2, age:40, hair:"blonde"}
1 row
216
From SQL to Cypher
14.4.Return
This is SQLs SELECT. We just put it in the end because it felt better to have it there you do a lot of
matching and filtering, and finally, you return something.
Aggregate queries work just like they do in SQL, apart from the fact that there is no explicit GROUP BY
clause. Everything in the return clause that is not an aggregate function will be used as the grouping
columns.
SQL Query.
SELECT "Person".name, count(*)
FROM "Person"
GROUP BY "Person".name
ORDER BY "Person".name
NAME
C2
Anakin
Bridget
2 rows
Cypher Query.
RETURN person.name, count(*)
ORDER BY person.name
person.name
count(*)
"Anakin"
"Bridget"
2 rows
Order by is the same in both languages ORDER BY expression ASC/DESC. Nothing weird here.
217
PartIV.Reference
The reference part is the authoritative source for details on Neo4j usage. It covers details on capabilities,
transactions, indexing and queries among other topics.
Chapter15.Capabilities
219
Capabilities
15.1.Data Security
Some data may need to be protected from unauthorized access (e.g., theft, modification). Neo4j does
not deal with data encryption explicitly, but supports all means built into the Java programming
language and the JVM to protect data by encrypting it before storing.
Furthermore, data can be easily secured by running on an encrypted datastore at the file system level.
Finally, data protection should be considered in the upper layers of the surrounding system in order to
prevent problems with scraping, malicious data insertion, and other threats.
220
Capabilities
15.2.Data Integrity
In order to keep data consistent, there needs to be mechanisms and structures that guarantee the
integrity of all stored data. In Neo4j, data integrity is maintained for the core graph engine together
with other data sources - see below.
15.2.1.Core Graph Engine

In Neo4j, the whole data model is stored as a graph on disk and persisted as part of every committed
transaction. In the storage layer, Relationships, Nodes, and Properties have direct pointers to each
other. This maintains integrity without the need for data duplication between the different backend
store files.
15.2.2.Different Data Sources

In a number of scenarios, the core graph engine is combined with other systems in order to achieve
optimal performance for non-graph lookups. For example, Apache Lucene is frequently used as an
additional index system for text queries that would otherwise be very processing-intensive in the
graph layer.
To keep these external systems in synchronization with each other, Neo4j provides full Two Phase
Commit transaction management, with rollback support over all data sources. Thus, failed index
insertions into Lucene can be transparently rolled back in all data sources and thus keep data up-todate.
221
Capabilities
15.3.Data Integration
Most enterprises rely primarily on relational databases to store their data, but this may cause
performance limitations. In some of these cases, Neo4j can be used as an extension to supplement
search/lookup for faster decision making. However, in any situation where multiple data repositories
contain the same data, synchronization can be an issue.
In some applications, it is acceptable for the search platform to be slightly out of sync with the
relational database. In others, tight data integrity (eg., between Neo4j and RDBMS) is necessary.
Typically, this has to be addressed for data changing in real-time and for bulk data changes happening
in the RDBMS.
A few strategies for synchronizing integrated data follows.
15.3.1.Event-based Synchronization
In this scenario, all data stores, both RDBMS and Neo4j, are fed with domain-specific events via
an event bus. Thus, the data held in the different backends is not actually synchronized but rather
replicated.
15.3.2.Periodic Synchronization
Another viable scenario is the periodic export of the latest changes in the RDBMS to Neo4j via
some form of SQL query. This allows a small amount of latency in the synchronization, but has the
advantage of using the RDBMS as the master for all data purposes. The same process can be applied
with Neo4j as the master data source.
15.3.3.Periodic Full Export/Import of Data

Using the Batch Inserter tools for Neo4j, even large amounts of data can be imported into the database
in very short times. Thus, a full export from the RDBMS and import into Neo4j becomes possible. If
the propagation lag between the RDBMS and Neo4j is not a big issue, this is a very viable solution.
222
Capabilities
15.4.Availability and Reliability

Most mission-critical systems require the database subsystem to be accessible at all times. Neo4j
ensures availability and reliability through a few different strategies.
15.4.1.Operational Availability
In order not to create a single point of failure, Neo4j supports different approaches which provide
transparent fallback and/or recovery from failures.
Online backup (Cold spare)
In this approach, a single instance of the master database is used, with Online Backup enabled. In
case of a failure, the backup files can be mounted onto a new Neo4j instance and reintegrated into the
application.
Online Backup High Availability (Hot spare)
Here, a Neo4j "backup" instance listens to online transfers of changes from the master. In the event of
a failure of the master, the backup is already running and can directly take over the load.
High Availability cluster
This approach uses a cluster of database instances, with one (read/write) master and a number of
(read-only) slaves. Failing slaves can simply be restarted and brought back online. Alternatively, a
new slave may be added by cloning an existing one. Should the master instance fail, a new master will
be elected by the remaining cluster nodes.
15.4.2.Disaster Recovery/ Resiliency

In cases of a breakdown of major part of the IT infrastructure, there need to be mechanisms in place
that enable the fast recovery and regrouping of the remaining services and servers. In Neo4j, there are
different components that are suitable to be part of a disaster recovery strategy.
Prevention
Online Backup High Availability to other locations outside the current data center.
Online Backup to different file system locations: this is a simpler form of backup, applying changes
directly to backup files; it is thus more suited for local backup scenarios.
Neo4j High Availability cluster: a cluster of one write-master Neo4j server and a number of readslaves, getting transaction logs from the master. Write-master failover is handled by quorum
election among the read-slaves for a new master.
Detection
SNMP and JMX monitoring can be used for the Neo4j database.
Correction
Online Backup: A new Neo4j server can be started directly on the backed-up files and take over
new requests.
Neo4j High Availability cluster: A broken Neo4j read slave can be reinserted into the cluster,
getting the latest updates from the master. Alternatively, a new server can be inserted by copying an
existing server and applying the latest updates to it.
223
Capabilities
15.5.Capacity
15.5.1.File Sizes
Neo4j relies on Javas Non-blocking I/O subsystem for all file handling. Furthermore, while the
storage file layout is optimized for interconnected data, Neo4j does not require raw devices. Thus,
filesizes are only limited by the underlying operating systems capacity to handle large files.
Physically, there is no built-in limit of the file handling capacity in Neo4j.
Neo4j tries to memory-map as much of the underlying store files as possible. If the available RAM is
not sufficient to keep all data in RAM, Neo4j will use buffers in some cases, reallocating the memorymapped high-performance I/O windows to the regions with the most I/O activity dynamically. Thus,
ACID speed degrades gracefully as RAM becomes the limiting factor.
15.5.2.Read speed
Enterprises want to optimize the use of hardware to deliver the maximum business value from
available resources. Neo4js approach to reading data provides the best possible usage of all available
hardware resources. Neo4j does not block or lock any read operations; thus, there is no danger for
deadlocks in read operations and no need for read transactions. With a threaded read access to the
database, queries can be run simultaneously on as many processors as may be available. This provides
very good scale-up scenarios with bigger servers.
15.5.3.Write speed
Write speed is a consideration for many enterprise applications. However, there are two different
scenarios:
1. sustained continuous operation and
2. bulk access (e.g., backup, initial or batch loading).
To support the disparate requirements of these scenarios, Neo4j supports two modes of writing to the
storage layer.
In transactional, ACID-compliant normal operation, isolation level is maintained and read operations
can occur at the same time as the writing process. At every commit, the data is persisted to disk and
can be recovered to a consistent state upon system failures. This requires disk write access and a real
flushing of data. Thus, the write speed of Neo4j on a single server in continuous mode is limited
by the I/O capacity of the hardware. Consequently, the use of fast SSDs is highly recommended for
production scenarios.
Neo4j has a Batch Inserter that operates directly on the store files. This mode does not provide
transactional security, so it can only be used when there is a single write thread. Because data is
written sequentially, and never flushed to the logical logs, huge performance boosts are achieved. The
Batch Inserter is optimized for non-transactional bulk import of large amounts of data.
15.5.4.Data size
In Neo4j, data size is mainly limited by the address space of the primary keys for Nodes,
Relationships, Properties and RelationshipTypes. Currently, the address space is as follows:
nodes
relationships
235 ( 34 billion)
235 ( 34 billion)
224
Capabilities
properties
relationship types
236 to 238 depending on property types (maximum

274 billion, always at least 68 billion)
215 ( 32 000)
225
Chapter16.Transaction Management
In order to fully maintain data integrity and ensure good transactional behavior, Neo4j supports the
ACID properties:
atomicity: If any part of a transaction fails, the database state is left unchanged.
consistency: Any transaction will leave the database in a consistent state.
isolation: During a transaction, modified data cannot be accessed by other operations.
durability: The DBMS can always recover the results of a committed transaction.
Specifically:
All database operations that access the graph, indexes, or the schema must be performed in a
transaction.
The default isolation level is READ_COMMITTED.
Data retrieved by traversals is not protected from modification by other transactions.
Non-repeatable reads may occur (i.e., only write locks are acquired and held until the end of the
transaction).
One can manually acquire write locks on nodes and relationships to achieve higher level of
isolation (SERIALIZABLE).
Locks are acquired at the Node and Relationship level.
Deadlock detection is built into the core transaction management.
226
Transaction Management
16.1.Interaction cycle
All database operations that access the graph, indexes, or the schema must be performed in a
transaction. Transactions are thread confined and can be nested as flat nested transactions.
Flat nested transactions means that all nested transactions are added to the scope of the top level
transaction. A nested transaction can mark the top level transaction for rollback, meaning the entire
transaction will be rolled back. To only rollback changes made in a nested transaction is not possible.
The interaction cycle of working with transactions looks like this:
1.
2.
3.
4.
Begin a transaction.
Perform database operations.
Mark the transaction as successful or not.
Finish the transaction.
It is very important to finish each transaction. The transaction will not release the locks or memory
it has acquired until it has been finished. The idiomatic use of transactions in Neo4j is to use a tryfinally block, starting the transaction and then try to perform the write operations. The last operation
in the try block should mark the transaction as successful while the finally block should finish the
transaction. Finishing the transaction will perform commit or rollback depending on the success
status.
Caution
All modifications performed in a transaction are kept in memory. This means that very
large updates have to be split into several top level transactions to avoid running out of
memory. It must be a top level transaction since splitting up the work in many nested
transactions will just add all the work to the top level transaction.
In an environment that makes use of thread pooling other errors may occur when failing to finish a
transaction properly. Consider a leaked transaction that did not get finished properly. It will be tied
to a thread and when that thread gets scheduled to perform work starting a new (what looks to be a)
top level transaction it will actually be a nested transaction. If the leaked transaction state is marked
for rollback (which will happen if a deadlock was detected) no more work can be performed on that
transaction. Trying to do so will result in error on each call to a write operation.
227
16.2.Isolation levels
By default a read operation will read the last committed value unless a local modification within the
current transaction exist. The default isolation level is very similar to READ_COMMITTED: reads do not
block or take any locks so non-repeatable reads can occur. It is possible to achieve a stronger isolation
level (such as REPETABLE_READ and SERIALIZABLE) by manually acquiring read and write locks.
228
16.3.Default locking behavior

When adding, changing or removing a property on a node or relationship a write lock will be taken
on the specific node or relationship.
When creating or deleting a node a write lock will be taken for the specific node.
When creating or deleting a relationship a write lock will be taken on the specific relationship and
both its nodes.
The locks will be added to the transaction and released when the transaction finishes.
229
16.4.Deadlocks
Since locks are used it is possible for deadlocks to happen. Neo4j will however detect any deadlock
(caused by acquiring a lock) before they happen and throw an exception. Before the exception is
thrown the transaction is marked for rollback. All locks acquired by the transaction are still being held
but will be released when the transaction is finished (in the finally block as pointed out earlier). Once
the locks are released other transactions that were waiting for locks held by the transaction causing the
deadlock can proceed. The work performed by the transaction causing the deadlock can then be retried
by the user if needed.
Experiencing frequent deadlocks is an indication of concurrent write requests happening in such a
way that it is not possible to execute them while at the same time live up to the intended isolation
and consistency. The solution is to make sure concurrent updates happen in a reasonable way. For
example given two specific nodes (A and B), adding or deleting relationships to both these nodes in
random order for each transaction will result in deadlocks when there are two or more transactions
doing that concurrently. One solution is to make sure that updates always happens in the same order
(first A then B). Another solution is to make sure that each thread/transaction does not have any
conflicting writes to a node or relationship as some other concurrent transaction. This can for example
be achieved by letting a single thread do all updates of a specific type.
Important
Deadlocks caused by the use of other synchronization than the locks managed by
Neo4j can still happen. Since all operations in the Neo4j API are thread safe unless
specified otherwise, there is no need for external synchronization. Other code that requires
synchronization should be synchronized in such a way that it never performs any Neo4j
operation in the synchronized block.
230
16.5.Delete semantics
When deleting a node or a relationship all properties for that entity will be automatically removed but
the relationships of a node will not be removed.
Caution
Neo4j enforces a constraint (upon commit) that all relationships must have a valid
start node and end node. In effect this means that trying to delete a node that still has
relationships attached to it will throw an exception upon commit. It is however possible
to choose in which order to delete the node and the attached relationships as long as no
relationships exist when the transaction is committed.
The delete semantics can be summarized in the following bullets:
All properties of a node or relationship will be removed when it is deleted.
A deleted node can not have any attached relationships when the transaction commits.
It is possible to acquire a reference to a deleted relationship or node that has not yet been
committed.
Any write operation on a node or relationship after it has been deleted (but not yet committed) will
throw an exception
After commit trying to acquire a new or work with an old reference to a deleted node or relationship
will throw an exception.
231
16.6.Creating unique nodes

In many use cases, a certain level of uniqueness is desired among entities. You could for instance
imagine that only one user with a certain e-mail address may exist in a system. If multiple concurrent
threads naively try to create the user, duplicates will be created. There are three main strategies for
ensuring uniqueness, and they all work across High Availability and single-instance deployments.
16.6.1.Single thread
By using a single thread, no two threads will even try to create a particular entity simultaneously. On
High Availability, an external single-threaded client can perform the operations on the cluster.
16.6.2.Get or create
The preferred way to get or create a unique node is to use unique constraints and Cypher. See
Section32.11.1, Get or create unique node using Cypher and unique constraints for more
information.
By using put-if-absent <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/index/
Index.html#putIfAbsent%28T,%20java.lang.String,%20java.lang.Object%29> functionality, entity
uniqueness can be guaranteed using a legacy index. Here the legacy index acts as the lock and will
only lock the smallest part needed to guaranteed uniqueness across threads and transactions.
See Section32.11.2, Get or create unique node using a legacy index for how to do this using the
core Java API. When using the REST API, see Section19.19, Unique Indexing.
16.6.3.Pessimistic locking
Important
While this is a working solution, please consider using the preferred Section16.6.2, Get
or create instead.
By using explicit, pessimistic locking, unique creation of entities can be achieved in a multi-threaded
environment. It is most commonly done by locking on a single or a set of common nodes.
See Section32.11.3, Pessimistic locking for node creation for how to do this using the core Java
API.
232
16.7.Transaction events
Transaction event handlers can be registered to receive Neo4j Transaction events. Once it has been
registered at a GraphDatabaseService instance it will receive events about what has happened in
each transaction which is about to be committed. Handlers wont get notified about transactions
which havent performed any write operation or wont be committed (either if Transactionsuccess()
hasnt been called or the transaction has been marked as failed Transactionfailure(). Right before
a transaction is about to be committed the beforeCommit method is called with the entire diff of
modifications made in the transaction. At this point the transaction is still running so changes can still
be made. However theres no guarantee that other handlers will see such changes since the order in
which handlers are executed is undefined. This method can also throw an exception and will, in such
a case, prevent the transaction from being committed (where a call to afterRollback will follow). If
beforeCommit is successfully executed the transaction will be committed and the afterCommit method
will be called with the same transaction data as well as the object returned from beforeCommit. This
assumes that all other handlers (if more were registered) also executed beforeCommit successfully.
233
Chapter17.Data Import
For importing data into Neo4j, see http://www.neo4j.org/develop/import.
For high-performance data import, see Chapter35, Batch Insertion.
234
Chapter18.Graph Algorithms
Neo4j graph algorithms is a component that contains Neo4j implementations of some common
algorithms for graphs. It includes algorithms like:
Shortest paths,
all paths,
all simple paths,
Dijkstra and
A*.
235
Graph Algorithms
18.1.Introduction
The graph algorithms are found in the neo4j-graph-algo component, which is included in the standard
Neo4j download.
Javadocs <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphalgo/packagesummary.html>
Download <http://search.maven.org/#search%7Cgav%7C1%7Cg%3A%22org.neo4j%22%20AND
%20a%3A%22neo4j-graph-algo%22>
Source code <https://github.com/neo4j/neo4j/tree/2.0.0/community/graph-algo>
For examples, see Section19.16, Graph Algorithms (REST API) and Section32.9, Graph
Algorithm examples (embedded database).
For information on how to use neo4j-graph-algo as a dependency with Maven and other dependency
management tools, see org.neo4j:neo4j-graph-algo <http://search.maven.org/search%7Cgav%7C1%7Cg
%3A%22org.neo4j%22%20AND%20a%3A%22neo4j-graph-algo%22> Note that it should be used with the same
version of org.neo4j:neo4j-kernel <http://search.maven.org/search%7Cgav%7C1%7Cg%3A%22org.neo4j
%22%20AND%20a%3A%22neo4j-kernel%22>. Different versions of the graph-algo and kernel components are
not compatible in the general case. Both components are included transitively by the org.neo4j:neo4j
<http://search.maven.org/search%7Cgav%7C1%7Cg%3A%22org.neo4j%22%20AND%20a%3A%22neo4j%22> artifact
which makes it simple to keep the versions in sync.
The starting point to find and use graph algorithms is GraphAlgoFactory <http://components.neo4j.org/
neo4j/2.0.0/apidocs/org/neo4j/graphalgo/GraphAlgoFactory.html> when using the Java Core API.
236
Chapter19.REST API
The Neo4j REST API is designed with discoverability in mind, so that you can start with a GET on the
Section19.3, Service root and from there discover URIs to perform other requests. The examples
below uses URIs in the examples; they are subject to change in the future, so for future-proofness
discover URIs where possible, instead of relying on the current layout. The default representation is
json <http://www.json.org/>, both for responses and for data sent with POST/PUT requests.
Below follows a listing of ways to interact with the REST API. For language bindings to the REST
API, see Chapter6, Languages.
To interact with the JSON interface you must explicitly set the request header Accept:application/json
for those requests that responds with data. You should also set the header Content-Type:application/
json if your request sends data, for example when youre creating a relationship. The examples include
the relevant request and response headers.
The server supports streaming results, with better performance and lower memory overhead. See
Section19.4, Streaming for more information.
237
REST API
19.1.Transactional HTTP endpoint
The Neo4j transactional HTTP endpoint allows you to execute a series of Cypher statements within
the scope of a transaction. The transaction may be kept open across multiple HTTP requests, until
the client chooses to commit or roll back. Each HTTP request can include a list of statements, and for
convenience you can include statements along with a request to begin or commit a transaction.
The server guards against orphaned transactions by using a timeout. If there are no requests for a
given transaction within the timeout period, the server will roll it back. You can configure the timeout
in the server configuration, by setting org.neo4j.server.transaction.timeout to the number of seconds
before timeout. The default timeout is 60 seconds.
The key difference between the transactional HTTP endpoint and the Cypher endpoint (see
Section19.5, Cypher queries via REST) is the ability to use the same transaction across multiple
HTTP requests. The cypher endpoint always attempts to commit a transaction at the end of each
HTTP request.
Note
The serialization format for cypher results is mostly the same as the cypher endpoint.
However, the format for raw entities is slightly less verbose and does not include
hypermedia links.
Note
Open transactions are not shared among members of an HA cluster. Therefore, if you use
this endpoint in an HA cluster, you must ensure that all requests for a given transaction are
sent to the same Neo4j instance.
Tip
In order to speed up queries in repeated scenarios, try not to use literals but replace them
with parameters wherever possible in order to let the server cache query plans.
19.1.1.Begin a transaction
You begin a new transaction by posting zero or more Cypher statements to the transaction endpoint.
The server will respond with the result of your statements, as well as the location of your open
transaction.
Example request
POST http://localhost:7474/db/data/transaction
Accept: application/json; charset=UTF-8
Content-Type: application/json
{
"statements" : [ {
"statement" : "CREATE (n {props}) RETURN n",
"parameters" : {
"props" : {
"name" : "My Node"
}
}
} ]
}
Example response
201: Created
238
REST API
Location: http://localhost:7474/db/data/transaction/7
{
"commit" : "http://localhost:7474/db/data/transaction/7/commit",
"results" : [ {
"columns" : [ "n" ],
"data" : [ {
"row" : [ {
"name" : "My Node"
} ]
} ]
} ],
"transaction" : {
"expires" : "Sun, 08 Dec 2013 15:05:52 +0000"
},
"errors" : [ ]
}
19.1.2.Execute statements in an open transaction

Given that you have an open transaction, you can make a number of requests, each of which executes
additional statements, and keeps the transaction open by resetting the transaction timeout.
Example request
POST http://localhost:7474/db/data/transaction/9
{
"statements" : [ {
"statement" : "CREATE n RETURN n"
} ]
}
Example response
200: OK
{
"results" : [ {
"data" : [ {
"row" : [ {
} ]
} ]
} ],
"transaction" : {
"expires" : "Sun, 08 Dec 2013 15:05:52 +0000"
},
"errors" : [ ]
}
19.1.3.Execute statements in an open transaction in REST format for the

return
Given that you have an open transaction, you can make a number of requests, each of which executes
additional statements, and keeps the transaction open by resetting the transaction timeout. Specifying
239
REST API
the REST format will give back full Neo4j Rest API representations of the Neo4j Nodes, Relationships
and Paths, if returned.
Example request
{
"statements" : [ {
"statement" : "CREATE n RETURN n",
"resultDataContents" : [ "REST" ]
} ]
}
Example response
200: OK
{
"results" : [ {
"data" : [ {
"rest" : [ {
"paged_traverse" : "http://localhost:7474/db/data/node/12/paged/traverse/{returnType}{?pageSize,leaseTime}",
"labels" : "http://localhost:7474/db/data/node/12/labels",
"create_relationship" : "http://localhost:7474/db/data/node/12/relationships",
"data" : {
}
} ]
} ]
} ],
"transaction" : {
"expires" : "Sun, 08 Dec 2013 15:05:48 +0000"
},
"errors" : [ ]
}
19.1.4.Reset transaction timeout of an open transaction

Every orphaned transaction is automatically expired after a period of inactivity. This may be
prevented by resetting the transaction timeout.
The timeout may be reset by sending a keep-alive request to the server that executes an empty list
of statements. This request will reset the transaction timeout and return the new time at which the
transaction will expire as an RFC1123 formatted timestamp value in the transaction section of the
response.
240
REST API
Example request
{
"statements" : [ ]
}
Example response
200: OK
{
"results" : [ ],
"transaction" : {
"expires" : "Sun, 08 Dec 2013 15:05:51 +0000"
},
"errors" : [ ]
}
19.1.5.Commit an open transaction

Given you have an open transaction, you can send a commit request. Optionally, you submit
additional statements along with the request that will be executed before committing the transaction.
Example request
POST http://localhost:7474/db/data/transaction/4/commit
{
"statements" : [ {
"statement" : "CREATE n RETURN id(n)"
} ]
}
Example response
200: OK
{
"results" : [ {
"columns" : [ "id(n)" ],
"data" : [ {
"row" : [ 14 ]
} ]
} ],
"errors" : [ ]
}
19.1.6.Rollback an open transaction

Given that you have an open transaction, you can send a roll back request. The server will roll back
the transaction.
241
REST API
Example request
DELETE http://localhost:7474/db/data/transaction/3
Example response
200: OK
Content-Type: application/json; charset=UTF-8
{
"results" : [ ],
"errors" : [ ]
}
19.1.7.Begin and commit a transaction in one request

If there is no need to keep a transaction open across multiple HTTP requests, you can begin a
transaction, execute statements, and commit with just a single HTTP request.
Example request
POST http://localhost:7474/db/data/transaction/commit
{
"statements" : [ {
"statement" : "CREATE n RETURN id(n)"
} ]
}
Example response
200: OK
{
"results" : [ {
"columns" : [ "id(n)" ],
"data" : [ {
"row" : [ 15 ]
} ]
} ],
"errors" : [ ]
}
19.1.8.Return results in graph format

If you want to understand the graph structure of nodes and relationships returned by your query, you
can specify the "graph" results data format. For example, this is useful when you want to visualise the
graph structure. The format collates all the nodes and relationships from all columns of the result, and
also flattens collections of nodes and relationships, including paths.
Example request
POST http://localhost:7474/db/data/transaction/commit
242
REST API
{
"statements" : [ {
"statement" : "CREATE ( bike:Bike { weight: 10 } )CREATE ( frontWheel:Wheel { spokes: 3 } )CREATE ( backWheel:Wheel { spokes: 32
"resultDataContents" : [ "row", "graph" ]
} ]
}
Example response
200: OK
{
"results" : [ {
"columns" : [ "bike", "p1", "p2" ],
"data" : [ {
"row" : [ {
"weight" : 10
}, [ {
"weight" : 10
}, {
"position" : 1
}, {
"spokes" : 3
} ], [ {
"weight" : 10
}, {
"position" : 2
}, {
"spokes" : 32
} ] ],
"graph" : {
"nodes" : [ {
"id" : "17",
"labels" : [ "Wheel" ],
"properties" : {
"spokes" : 3
}
}, {
"id" : "16",
"labels" : [ "Bike" ],
"properties" : {
"weight" : 10
}
}, {
"id" : "18",
"labels" : [ "Wheel" ],
"properties" : {
"spokes" : 32
}
} ],
"relationships" : [ {
"id" : "9",
"type" : "HAS",
"startNode" : "16",
"endNode" : "17",
"properties" : {
"position" : 1
}
}, {
243
REST API
"id" : "10",
"type" : "HAS",
"startNode" : "16",
"endNode" : "18",
"properties" : {
"position" : 2
}
} ]
}
} ]
} ],
"errors" : [ ]
}
19.1.9.Handling errors
The result of any request against the transaction endpoint is streamed back to the client. Therefore
the server does not know whether the request will be successful or not when it sends the HTTP status
code.
Because of this, all requests against the transactional endpoint will return 200 or 201 status code,
regardless of whether statements were successfully executed. At the end of the response payload,
the server includes a list of errors that occurred while executing statements. If this list is empty, the
request completed successfully.
If any errors occur while executing statements, the server will roll back the transaction.
In this example, we send the server an invalid statement to demonstrate error handling.
For more information on the status codes, see Section19.2, Neo4j Status Codes.
Example request
POST http://localhost:7474/db/data/transaction/8/commit
{
"statements" : [ {
"statement" : "This is not a valid Cypher Statement."
} ]
}
Example response
200: OK
{
"results" : [ ],
"errors" : [ {
"code" : "Neo.ClientError.Statement.InvalidSyntax",
"message" : "Invalid input 'T': expected SingleStatement (line 1, column 1)\n\"This is not a valid Cypher Statement.\"\n ^"
} ]
}
244
REST API
19.2.Neo4j Status Codes
The transactional endpoint may in any response include zero or more status codes, indicating issues or
information for the client. Each status code follows the same format: "Neo.[Classification].[Category].
[Title]". The fact that a status code is returned by the server does always mean there is a fatal error.
Status codes can also indicate transient problems that may go away if you retry the request.
What the effect of the status code is can be determined by its classification.
Note
This is not the same thing as HTTP status codes. Neo4j Status Codes are returned in the
response body, at the very end of the response.
19.2.1.Classifications
Classification
Description
Effect on
transaction
ClientError
The Client sent a bad request - changing the request might

yield a successful outcome.
None
DatabaseError
The database failed to service the request.
Rollback
TransientError
The database cannot service the request right now, retrying

later might yield a successful outcome.
None
19.2.2.Status codes
This is a complete list of all status codes Neo4j may return, and what they mean.
Status Code
Description
Neo. ClientError. Request. Invalid
The client provided an invalid request.
Neo. ClientError. Request. InvalidFormat
The client provided a request that was missing

required fields, or had values that are not allowed.
Neo. ClientError. Schema. ConstraintViolation
A constraint imposed by the database was

violated.
Neo. ClientError. Schema. NoSuchConstraint
The request (directly or indirectly) referred to a

constraint that does not exist.
Neo. ClientError. Schema. NoSuchIndex
The request (directly or indirectly) referred to an

index that does not exist.
Neo. ClientError. Statement. ArithmeticError
Invalid use of arithmetic, such as dividing by

zero.
Neo. ClientError. Statement. ConstraintViolation
A constraint imposed by the statement is violated

by the data in the database.
Neo. ClientError. Statement. EntityNotFound
The statement is directly referring to an entity that

does not exist.
Neo. ClientError. Statement. InvalidSemantics
The statement is syntactically valid, but expresses

something that the database cannot do.
Neo. ClientError. Statement. InvalidSyntax
The statement contains invalid or unsupported

syntax.
245
REST API
Status Code
Description
Neo. ClientError. Statement. InvalidType
The statement is attempting to perform operations

on values with types that are not supported by the
operation.
Neo. ClientError. Statement. ParameterMissing
The statement is referring to a parameter that was

not provided in the request.
Neo. ClientError. Transaction. ConcurrentRequest
There were concurrent requests accessing the

same transaction, which is not allowed.
Neo. ClientError. Transaction. UnknownId
The request referred to a transaction that does not

exist.
Neo. DatabaseError. General. FailedIndex
The request (directly or indirectly) referred to an

index that is in a failed state. The index needs to
be dropped and recreated manually.
Neo. DatabaseError. General. UnknownFailure
An unknown failure occurred.
Neo. DatabaseError. Statement. ExecutionFailure
The database was unable to execute the statement.
Neo. DatabaseError. Transaction. CouldNotBegin
The database was unable to start the transaction.
Neo. DatabaseError. Transaction. CouldNotCommit
The database was unable to commit the

transaction.
Neo. DatabaseError. Transaction. CouldNotRollback
The database was unable to roll back the

transaction.
Neo. TransientError. Network. UnknownFailure
An unknown network failure occurred, a retry

may resolve the issue.
246
REST API
19.3.Service root
19.3.1.Get service root
The service root is your starting point to discover the REST API. It contains the basic starting points
for the database, and some version and extension information.
Figure19.1.Final Graph
Example request
GET http://localhost:7474/db/data/
Example response
200: OK
{
"extensions" : {
},
"node" : "http://localhost:7474/db/data/node",
"node_index" : "http://localhost:7474/db/data/index/node",
"relationship_index" : "http://localhost:7474/db/data/index/relationship",
"extensions_info" : "http://localhost:7474/db/data/ext",
"relationship_types" : "http://localhost:7474/db/data/relationship/types",
"batch" : "http://localhost:7474/db/data/batch",
"cypher" : "http://localhost:7474/db/data/cypher",
"transaction" : "http://localhost:7474/db/data/transaction",
"neo4j_version" : "2.0.0"
}
247
REST API
19.4.Streaming
All responses from the REST API can be transmitted as JSON streams, resulting in better performance
and lower memory overhead on the server side. To use streaming, supply the header X-Stream: true
with each request.
Example request
GET http://localhost:7474/db/data/
Accept: application/json
X-Stream: true
Example response
200: OK
Content-Type: application/json; charset=UTF-8; stream=true
{
"extensions" : {
},
"relationship_types" : "http://localhost:7474/db/data/relationship/types",
"batch" : "http://localhost:7474/db/data/batch",
"cypher" : "http://localhost:7474/db/data/cypher",
"transaction" : "http://localhost:7474/db/data/transaction",
"neo4j_version" : "2.0.0"
}
248
REST API
19.5.Cypher queries via REST

The Neo4j REST API allows querying with Cypher, see PartIII, Cypher Query Language. The
results are returned as a list of string headers (columns), and a data part, consisting of a list of all rows,
every row consisting of a list of REST representations of the field value Node, Relationship, Path or
any simple value like String.
Tip
In order to speed up queries in repeated scenarios, try not to use literals but replace
them with parameters wherever possible in order to let the server cache query plans, see
Section19.5.1, Use parameters for details. Also see Section7.5, Parameters for where
parameters can be used.
19.5.1.Use parameters
Cypher supports queries with parameters which are submitted as JSON.
MATCH (x { name: { startName }})-[r]-(friend)
WHERE friend.name = { name }
RETURN TYPE(r)
Node[ 61] nam e = 'I'
know
Node[ 60] nam e = 'you'
Example request
POST http://localhost:7474/db/data/cypher
{
"query" : "MATCH (x {name: {startName}})-[r]-(friend) WHERE friend.name = {name} RETURN TYPE(r)",
"params" : {
"startName" : "I",
"name" : "you"
}
}
Example response
200: OK
{
"columns" : [ "TYPE(r)" ],
"data" : [ [ "know" ] ]
}
249
REST API
19.5.2.Create a node
Create a node with a label and a property using Cypher. See the request for the parameter sent with the
query.
CREATE (n:Person { name : { name }})
RETURN n
Node[ 47] : Person
nam e = 'Andres'
Example request
{
"query" : "CREATE (n:Person { name : {name} }) RETURN n",
"params" : {
"name" : "Andres"
}
}
Example response
200: OK
{
"data" : [ [ {
"data" : {
"name" : "Andres"
},
"extensions" : {
},
"incoming_typed_relationships" : "http://localhost:7474/db/data/node/47/relationships/in/{-list|&|types}"
} ] ]
}
19.5.3.Create a node with multiple properties

Create a node with a label and multiple properties using Cypher. See the request for the parameter sent
with the query.
250
REST API
RETURN n
Node[ 44] : Person
nam e = 'Michael'
posit ion = 'Developer'
awesom e = t rue
children = 3
Example request
{
"query" : "CREATE (n:Person { props } ) RETURN n",
"params" : {
"props" : {
"name" : "Michael",
"awesome" : true,
"children" : 3
}
}
}
Example response
200: OK
{
"data" : [ [ {
"data" : {
"awesome" : true,
"name" : "Michael",
"children" : 3
},
"extensions" : {
},
} ] ]
}
251
REST API
19.5.4.Create mutiple nodes with properties

Create multiple nodes with properties using Cypher. See the request for the parameter sent with the
query.
RETURN n
Node[ 48] : Person
Node[ 49] : Person
nam e = 'Andres'
nam e = 'Michael'
Example request
{
"query" : "CREATE (n:Person { props } ) RETURN n",
"params" : {
"props" : [ {
"name" : "Andres",
}, {
"name" : "Michael",
} ]
}
}
Example response
200: OK
{
"data" : [ [ {
"data" : {
"name" : "Andres"
},
"extensions" : {
},
252
REST API
} ], [ {
"data" : {
"name" : "Michael"
},
"extensions" : {
},
} ] ]
}
19.5.5.Set all properties on a node using Cypher

Set all properties on a node.
CREATE (n:Person { name: 'this property is to be deleted' })
SET n = { props }
RETURN n
Node[ 75] : Person
awesom e = t rue
children = 3
first Nam e = 'Michael'
Example request
{
"query" : "CREATE (n:Person { name: 'this property is to be deleted' } ) SET n = { props } RETURN n",
"params" : {
"props" : {
"firstName" : "Michael",
"awesome" : true,
"children" : 3
}
}
}
Example response
200: OK
253
REST API
{
"data" : [ [ {
"data" : {
"awesome" : true,
"children" : 3,
"firstName" : "Michael"
},
"extensions" : {
},
} ] ]
}
19.5.6.Send a query
A simple query returning all nodes connected to some node, returning the node and the name property,
if it exists, otherwise NULL:
MATCH (x { name: 'I' })-[r]->(n)
RETURN type(r), n.name, n.age
know
know
Node[ 68] age = 25

nam e = 'him '
Example request
{
"query" : "MATCH (x {name: 'I'})-[r]->(n) RETURN type(r), n.name, n.age",
"params" : {
}
}
Example response
254
REST API
200: OK
{
"columns" : [ "type(r)", "n.name", "n.age" ],
"data" : [ [ "know", "him", 25 ], [ "know", "you", null ] ]
}
19.5.7.Return paths
Paths can be returned just like other return types.
MATCH path =(x { name: 'I' })--(friend)
RETURN path, friend.name
know
Example request
{
"query" : "MATCH path = (x {name: 'I'})--(friend) RETURN path, friend.name",
"params" : {
}
}
Example response
200: OK
{
"columns" : [ "path", "friend.name" ],
"data" : [ [ {
"start" : "http://localhost:7474/db/data/node/74",
"nodes" : [ "http://localhost:7474/db/data/node/74", "http://localhost:7474/db/data/node/73" ],
"length" : 1,
"relationships" : [ "http://localhost:7474/db/data/relationship/20" ],
"end" : "http://localhost:7474/db/data/node/73"
}, "you" ] ]
}
19.5.8.Nested results
When sending queries that return nested results like list and maps, these will get serialized into nested
JSON representations according to their types.
255
REST API
MATCH (n)
WHERE n.name IN ['I', 'you']
RETURN collect(n.name)
know
Example request
{
"query" : "MATCH (n) WHERE n.name in ['I', 'you'] RETURN collect(n.name)",
"params" : {
}
}
Example response
200: OK
{
"columns" : [ "collect(n.name)" ],
"data" : [ [ [ "you", "I" ] ] ]
}
19.5.9.Retrieve query metadata

By passing in an additional GET parameter when you execute Cypher queries, metadata about the
query will be returned, such as how many labels were added or removed by the query.
MATCH (n { name: 'I' })
SET n:Actor
REMOVE n:Director
RETURN labels(n)
Node[ 72] : Act or
nam e = 'I'
Example request
POST http://localhost:7474/db/data/cypher?includeStats=true
256
REST API
{
"query" : "MATCH (n {name: 'I'}) SET n:Actor REMOVE n:Director RETURN labels(n)",
"params" : {
}
}
Example response
200: OK
{
"columns" : [ "labels(n)" ],
"data" : [ [ [ "Actor" ] ] ],
"stats" : {
"relationships_created" : 0,
"nodes_deleted" : 0,
"relationship_deleted" : 0,
"indexes_added" : 0,
"properties_set" : 0,
"constraints_removed" : 0,
"indexes_removed" : 0,
"labels_removed" : 1,
"constraints_added" : 0,
"labels_added" : 1,
"nodes_created" : 0,
"contains_updates" : true
}
}
19.5.10.Errors
Errors on the server will be reported as a JSON-formatted message, exception name and stacktrace.
MATCH (x { name: 'I' })
RETURN x.dummy/0
know
Example request
{
"query" : "MATCH (x {name: 'I'}) RETURN x.dummy/0",
257
REST API
"params" : {
}
}
Example response
400: Bad Request
{
"message": "/ by zero",

"exception": "ArithmeticException",
"fullname": "org.neo4j.cypher.ArithmeticException",
"stacktrace": [
"org.neo4j.cypher.internal.compiler.v2_0.commands.expressions.Divide.apply(Divide.scala:36)",
"org.neo4j.cypher.internal.compiler.v2_0.pipes.ExtractPipe$$anonfun$internalCreateResults$1$$anonfun$apply$2.apply(ExtractPipe.sc
"org.neo4j.cypher.internal.compiler.v2_0.pipes.ExtractPipe$$anonfun$internalCreateResults$1$$anonfun$apply$2.apply(ExtractPipe.sc
"scala.collection.immutable.Map$Map1.foreach(Map.scala:109)",
"org.neo4j.cypher.internal.compiler.v2_0.pipes.ExtractPipe$$anonfun$internalCreateResults$1.apply(ExtractPipe.scala:49)",
"org.neo4j.cypher.internal.compiler.v2_0.pipes.ExtractPipe$$anonfun$internalCreateResults$1.apply(ExtractPipe.scala:48)",
"scala.collection.Iterator$$anon$11.next(Iterator.scala:328)",
"org.neo4j.cypher.internal.compiler.v2_0.ClosingIterator$$anonfun$next$1.apply(ClosingIterator.scala:46)",
"org.neo4j.cypher.internal.compiler.v2_0.ClosingIterator$$anonfun$next$1.apply(ClosingIterator.scala:45)",
"org.neo4j.cypher.internal.compiler.v2_0.ClosingIterator.failIfThrows(ClosingIterator.scala:91)",
"org.neo4j.cypher.internal.compiler.v2_0.ClosingIterator.next(ClosingIterator.scala:45)",
"org.neo4j.cypher.internal.compiler.v2_0.PipeExecutionResult.next(PipeExecutionResult.scala:168)",
"org.neo4j.cypher.internal.compiler.v2_0.PipeExecutionResult.next(PipeExecutionResult.scala:34)",
"scala.collection.convert.Wrappers$IteratorWrapper.next(Wrappers.scala:30)",
"org.neo4j.cypher.internal.compiler.v2_0.PipeExecutionResult$$anon$1.next(PipeExecutionResult.scala:76)",
"org.neo4j.helpers.collection.ExceptionHandlingIterable$1.next(ExceptionHandlingIterable.java:53)",
"org.neo4j.helpers.collection.IteratorWrapper.next(IteratorWrapper.java:47)",
"org.neo4j.server.rest.repr.ListRepresentation.serialize(ListRepresentation.java:64)",
"org.neo4j.server.rest.repr.Serializer.serialize(Serializer.java:75)",
"org.neo4j.server.rest.repr.MappingSerializer.putList(MappingSerializer.java:61)",
"org.neo4j.server.rest.repr.CypherResultRepresentation.serialize(CypherResultRepresentation.java:83)",
"org.neo4j.server.rest.repr.MappingRepresentation.serialize(MappingRepresentation.java:41)",
"org.neo4j.server.rest.repr.OutputFormat.assemble(OutputFormat.java:215)",
"org.neo4j.server.rest.repr.OutputFormat.formatRepresentation(OutputFormat.java:147)",
"org.neo4j.server.rest.repr.OutputFormat.response(OutputFormat.java:130)",
"org.neo4j.server.rest.repr.OutputFormat.ok(OutputFormat.java:67)",
"org.neo4j.server.rest.web.CypherService.cypher(CypherService.java:101)",
"java.lang.reflect.Method.invoke(Method.java:606)",
"org.neo4j.server.rest.transactional.TransactionalRequestDispatcher.dispatch(TransactionalRequestDispatcher.java:132)",
"org.neo4j.server.rest.security.SecurityFilter.doFilter(SecurityFilter.java:112)",
"java.lang.Thread.run(Thread.java:724)"
]
}
258
REST API
19.6.Property values
The REST API allows setting properties on nodes and relationships through direct RESTful
operations. However, there are restrictions as to what types of values can be used as property values.
Allowed value types are as follows:
Numbers: Both integer values, with capacity as Javas Long type, and floating points, with capacity
as Javas Double.
Booleans.
Strings.
Arrays of the basic types above.
19.6.1.Arrays
There are two important points to be made about array values. First, all values in the array must be of
the same type. That means either all integers, all floats, all booleans or all strings. Mixing types is not
currently supported.
Second, storing empty arrays is only possible given certain preconditions. Because the JSON transfer
format does not contain type information for arrays, type is inferred from the values in the array. If
the array is empty, the Neo4j Server cannot determine the type. In these cases, it will check if an array
is already stored for the given property, and will use the stored arrays type when storing the empty
array. If no array exists already, the server will reject the request.
19.6.2.Property keys
You can list all property keys ever used in the database. This includes and property keys you have
used, but deleted.
There is currently no way to tell which ones are in use and which ones are not, short of walking the
entire set of properties in the database.
19.6.3.List all property keys

Example request
GET http://localhost:7474/db/data/propertykeys
Example response
200: OK
[ "firstName", "some-key", "sequence", "awesome", "prop", "cost", "non-existent", "", "value", "key", "array1", "ID", "array
259
REST API
19.7.Nodes
19.7.1.Create node
Node[ 8]
Example request
POST http://localhost:7474/db/data/node
Example response
201: Created
Location: http://localhost:7474/db/data/node/8
{
"extensions" : {
},
"data" : {
}
}
19.7.2.Create node with properties

Node[ 4] foo = 'bar'
Example request
{
"foo" : "bar"
260
REST API
}
Example response
201: Created
Content-Length: 1156
Location: http://localhost:7474/db/data/node/4
{
"extensions" : {
},
"data" : {
"foo" : "bar"
}
}
19.7.3.Get node
Note that the response contains URI/templates for the available operations for getting properties and
relationships.
Node[ 30]
Example request
GET http://localhost:7474/db/data/node/30
Example response
200: OK
{
"extensions" : {
},
261
REST API
"data" : {
}
}
19.7.4.Get non-existent node

Node[ 34]
Example request
GET http://localhost:7474/db/data/node/3400000
Example response
404: Not Found
{
"message": "Cannot find node with id [3400000] in database.",
"exception": "NodeNotFoundException",
"fullname": "org.neo4j.server.rest.web.NodeNotFoundException",
"stacktrace": [
"org.neo4j.server.rest.web.DatabaseActions.node(DatabaseActions.java:183)",
"org.neo4j.server.rest.web.DatabaseActions.getNode(DatabaseActions.java:228)",
"org.neo4j.server.rest.web.RestfulGraphDatabase.getNode(RestfulGraphDatabase.java:266)",
]
}
19.7.5.Delete node
Example request
DELETE http://localhost:7474/db/data/node/5
Example response
204: No Content
262
REST API
19.7.6.Nodes with relationships cannot be deleted

The relationships on a node has to be deleted before the node can be deleted.
Node[ 12]
LOVES
Node[ 13]
Example request
DELETE http://localhost:7474/db/data/node/12
Example response
409: Conflict
{
"message": "The node with id 12 cannot be deleted. Check that the node is orphaned before deletion.",
"exception": "OperationFailureException",
"fullname": "org.neo4j.server.rest.web.OperationFailureException",
"stacktrace": [
"org.neo4j.server.rest.web.DatabaseActions.deleteNode(DatabaseActions.java:237)",
"org.neo4j.server.rest.web.RestfulGraphDatabase.deleteNode(RestfulGraphDatabase.java:280)",
]
}
263
REST API
19.8.Relationships
Relationships are a first class citizen in the Neo4j REST API. They can be accessed either stand-alone
or through the nodes they are attached to.
The general pattern to get relationships from a node is:
GET http://localhost:7474/db/data/node/123/relationships/{dir}/{-list|&|types}
Where dir is one of all, in, out and types is an ampersand-separated list of types. See the examples
below for more information.
19.8.1.Get Relationship by ID
know
Example request
GET http://localhost:7474/db/data/relationship/30
Example response
200: OK
{
"extensions" : {
},
"property" : "http://localhost:7474/db/data/relationship/30/properties/{key}",
"self" : "http://localhost:7474/db/data/relationship/30",
"properties" : "http://localhost:7474/db/data/relationship/30/properties",
"type" : "know",
"end" : "http://localhost:7474/db/data/node/138",
"data" : {
}
}
19.8.2.Create relationship
Upon successful creation of a relationship, the new relationship is returned.
264
REST API
knows LOVES
Example request
POST http://localhost:7474/db/data/node/1/relationships
{
"to" : "http://localhost:7474/db/data/node/0",
"type" : "LOVES"
}
Example response
201: Created
Location: http://localhost:7474/db/data/relationship/1
{
"extensions" : {
},
"type" : "LOVES",
"data" : {
}
}
19.8.3.Create a relationship with properties

Upon successful creation of a relationship, the new relationship is returned.
Figure19.21.Starting Graph
knows
265
REST API
knows
LOVES
foo = 'bar'
Example request
POST http://localhost:7474/db/data/node/11/relationships
{
"type" : "LOVES",
"data" : {
"foo" : "bar"
}
}
Example response
201: Created
Location: http://localhost:7474/db/data/relationship/8
{
"extensions" : {
},
"type" : "LOVES",
"data" : {
"foo" : "bar"
}
}
266
REST API
19.8.4.Delete relationship
Node[ 125] nam e = 'Rom eo'
LOVES
cost = 'high'
Node[ 124] nam e = 'Juliet '
Example request
DELETE http://localhost:7474/db/data/relationship/23
Example response
204: No Content
19.8.5.Get all properties on a relationship

LOVES
since = '1day'
cost = 'high'
Example request
GET http://localhost:7474/db/data/relationship/26/properties
Example response
267
REST API
200: OK
{
"cost" : "high",
"since" : "1day"
}
19.8.6.Set all properties on a relationship

LOVES
cost = 'high'
LOVES
happy = false
Example request
PUT http://localhost:7474/db/data/relationship/31/properties
{
"happy" : false
}
Example response
204: No Content
268
REST API
19.8.7.Get single property on a relationship

LOVES
cost = 'high'
Example request
GET http://localhost:7474/db/data/relationship/27/properties/cost
Example response
200: OK
"high"
19.8.8.Set single property on a relationship

LOVES
cost = 'high'
LOVES
cost = 'deadly'
269
REST API
Example request
PUT http://localhost:7474/db/data/relationship/25/properties/cost
"deadly"
Example response
204: No Content
19.8.9.Get all relationships

Node[ 301]
Node[ 303]
LIKES
Node[ 299]
LIKES
HATES
Node[ 300]
Node[ 302]
Example request
GET http://localhost:7474/db/data/node/299/relationships/all
Example response
200: OK
[ {
"data" : {
},
"type" : "LIKES",
"extensions" : {
},
}, {
270
REST API
"data" : {
},
"type" : "LIKES",
"extensions" : {
},
}, {
"data" : {
},
"type" : "HATES",
"extensions" : {
},
} ]
19.8.10.Get incoming relationships

Node[ 320]
Node[ 322]
LIKES
Node[ 318]
LIKES
HATES
Node[ 319]
Node[ 321]
Example request
GET http://localhost:7474/db/data/node/318/relationships/in
Example response
200: OK
[ {
"data" : {
},
271
REST API
"type" : "LIKES",
"extensions" : {
},
} ]
19.8.11.Get outgoing relationships

Node[ 345]
Node[ 347]
LIKES
Node[ 343]
LIKES
HATES
Node[ 344]
Node[ 346]
Example request
GET http://localhost:7474/db/data/node/343/relationships/out
Example response
200: OK
[ {
"data" : {
},
"type" : "LIKES",
"extensions" : {
},
}, {
"data" : {
},
"type" : "HATES",
"extensions" : {
272
REST API
},
} ]
19.8.12.Get typed relationships

Note that the "&" needs to be encoded like "%26" for example when using cURL <http://curl.haxx.se/>
from the terminal.
Node[ 276]
Node[ 278]
LIKES
Node[ 274]
LIKES
HATES
Node[ 275]
Node[ 277]
Example request
GET http://localhost:7474/db/data/node/274/relationships/all/LIKES&HATES
Example response
200: OK
[ {
"data" : {
},
"type" : "LIKES",
"extensions" : {
},
}, {
"data" : {
},
"type" : "LIKES",
"extensions" : {
},
273
REST API
}, {
"data" : {
},
"type" : "HATES",
"extensions" : {
},
} ]
19.8.13.Get relationships on a node without relationships

Node[ 330]
Node[ 332]
LIKES
Node[ 328]
LIKES
HATES
Node[ 329]
Node[ 331]
Example request
GET http://localhost:7474/db/data/node/332/relationships/all
Example response
200: OK
[ ]
274
REST API
19.9.Relationship types
19.9.1.Get relationship types
Example request
GET http://localhost:7474/db/data/relationship/types
Example response
200: OK
[ "know", "KNOWS", "LOVES" ]
275
REST API
19.10.Node properties
19.10.1.Set property on node
Setting different properties will retain the existing ones for this node. Note that a single value are
submitted not as a map but just as a value (which is valid JSON) like in the example below.
foo2 = 'bar2'
Example request
PUT http://localhost:7474/db/data/node/11/properties/foo
"bar"
Example response
204: No Content
19.10.2.Update node properties

This will replace all existing properties on the node with the new set of attributes.
Node[ 9] age = '18'
knows
Node[ 10] nam e = 'joe'
Example request
PUT http://localhost:7474/db/data/node/9/properties
{
"age" : "18"
}
Example response
204: No Content
276
REST API
19.10.3.Get properties for node

Example request
GET http://localhost:7474/db/data/node/36/properties
Example response
200: OK
{
"foo" : "bar"
}
19.10.4.Property values can not be null

This example shows the response you get when trying to set a property to null.
Example request
{
"foo" : null
}
Example response
400: Bad Request
{
"message": "Could not set property \"foo\", unsupported type: null",
"exception": "PropertyValueException",
"fullname": "org.neo4j.server.rest.web.PropertyValueException",
"stacktrace": [
"org.neo4j.server.rest.domain.PropertySettingStrategy.setProperty(PropertySettingStrategy.java:141)",
"org.neo4j.server.rest.domain.PropertySettingStrategy.setProperties(PropertySettingStrategy.java:88)",
"org.neo4j.server.rest.web.DatabaseActions.createNode(DatabaseActions.java:214)",
"org.neo4j.server.rest.web.RestfulGraphDatabase.createNode(RestfulGraphDatabase.java:239)",
]
}
277
REST API
19.10.5.Property values can not be nested

Nesting properties is not supported. You could for example store the nested JSON as a string instead.
Example request
POST http://localhost:7474/db/data/node/
{
"foo" : {
"bar" : "baz"
}
}
Example response
400: Bad Request
{
"message": "Could not set property \"foo\", unsupported type: {bar\u003dbaz}",
"exception": "PropertyValueException",
"fullname": "org.neo4j.server.rest.web.PropertyValueException",
"stacktrace": [
"org.neo4j.server.rest.domain.PropertySettingStrategy.setProperty(PropertySettingStrategy.java:141)",
"org.neo4j.server.rest.domain.PropertySettingStrategy.setProperties(PropertySettingStrategy.java:88)",
"org.neo4j.server.rest.web.DatabaseActions.createNode(DatabaseActions.java:214)",
"org.neo4j.server.rest.web.RestfulGraphDatabase.createNode(RestfulGraphDatabase.java:239)",
]
}
19.10.6.Delete all properties from node

Node[ 251]
Example request
DELETE http://localhost:7474/db/data/node/251/properties
Example response
204: No Content
19.10.7.Delete a named property from a node

To delete a single property from a node, see the example below.
278
REST API
Node[ 250] nam e = 't obias'
Node[ 250]
Example request
DELETE http://localhost:7474/db/data/node/250/properties/name
Example response
204: No Content
279
REST API
19.11.Relationship properties
19.11.1.Update relationship properties
Node[ 20]
KNOWS
jim = 't obias'
Node[ 21]
Example request
PUT http://localhost:7474/db/data/relationship/9/properties
{
"jim" : "tobias"
}
Example response
204: No Content
19.11.2.Remove properties from a relationship

LOVES
Example request
DELETE http://localhost:7474/db/data/relationship/21/properties
Example response
204: No Content
280
REST API
19.11.3.Remove property from a relationship

See the example request below.
LOVES
cost = 'high'
LOVES
Example request
DELETE http://localhost:7474/db/data/relationship/24/properties/cost
Example response
204: No Content
19.11.4.Remove non-existent property from a relationship

Attempting to remove a property that doesnt exist results in an error.
LOVES
cost = 'high'
281
REST API
Example request
DELETE http://localhost:7474/db/data/relationship/22/properties/non-existent
Example response
404: Not Found
{
"message": "Relationship[22] does not have a property \"non-existent\"",
"exception": "NoSuchPropertyException",
"fullname": "org.neo4j.server.rest.web.NoSuchPropertyException",
"stacktrace": [
"org.neo4j.server.rest.web.DatabaseActions.removeRelationshipProperty(DatabaseActions.java:657)",
"org.neo4j.server.rest.web.RestfulGraphDatabase.deleteRelationshipProperty(RestfulGraphDatabase.java:790)",
]
}
19.11.5.Remove properties from a non-existing relationship

Attempting to remove all properties from a relationship which doesnt exist results in an error.
know
Example request
DELETE http://localhost:7474/db/data/relationship/1234/properties
Example response
404: Not Found
{
"exception": "RelationshipNotFoundException",
"fullname": "org.neo4j.server.rest.web.RelationshipNotFoundException",
"stacktrace": [
"org.neo4j.server.rest.web.DatabaseActions.relationship(DatabaseActions.java:197)",
"org.neo4j.server.rest.web.DatabaseActions.removeAllRelationshipProperties(DatabaseActions.java:647)",
"org.neo4j.server.rest.web.RestfulGraphDatabase.deleteAllRelationshipProperties(RestfulGraphDatabase.java:769)",
282
REST API
]
}
19.11.6.Remove property from a non-existing relationship

Attempting to remove a property from a relationship which doesnt exist results in an error.
know
Example request
DELETE http://localhost:7474/db/data/relationship/1234/properties/cost
Example response
404: Not Found
{
"exception": "RelationshipNotFoundException",
"fullname": "org.neo4j.server.rest.web.RelationshipNotFoundException",
"stacktrace": [
"org.neo4j.server.rest.web.DatabaseActions.relationship(DatabaseActions.java:197)",
"org.neo4j.server.rest.web.DatabaseActions.removeRelationshipProperty(DatabaseActions.java:653)",
"org.neo4j.server.rest.web.RestfulGraphDatabase.deleteRelationshipProperty(RestfulGraphDatabase.java:790)",
]
}
283
REST API
19.12.Node labels
19.12.1.Adding a label to a node
Node[ 101] : Person
nam e = 'Clint East wood'
Example request
POST http://localhost:7474/db/data/node/101/labels
"Person"
Example response
204: No Content
19.12.2.Adding multiple labels to a node

Node[ 112] : Person, Act or
Example request
[ "Person", "Actor" ]
Example response
204: No Content
19.12.3.Adding a label with an invalid name

Labels with empty names are not allowed, however, all other valid strings are accepted as label
names. Adding an invalid label to a node will lead to a HTTP 400 response.
Node[ 119] nam e = 'Clint East wood'
284
REST API
Example request
""
Example response
400: Bad Request
{
"message": "Unable to add label, see nested exception.",

"exception": "BadInputException",
"fullname": "org.neo4j.server.rest.repr.BadInputException",
"stacktrace": [
"org.neo4j.server.rest.web.DatabaseActions.addLabelToNode(DatabaseActions.java:328)",
"org.neo4j.server.rest.web.RestfulGraphDatabase.addNodeLabel(RestfulGraphDatabase.java:440)",
],
"cause": {
"message": "Invalid label name \u0027\u0027.",
"cause": {
"message": "\u0027\u0027 is not a valid token name. Only non-null, non-empty strings are allowed.",
"exception": "IllegalTokenNameException",
"stacktrace": [
"org.neo4j.kernel.impl.api.DataIntegrityValidatingStatementOperations.checkValidTokenName(DataIntegrityValidatingStatementOpe
"org.neo4j.kernel.impl.api.DataIntegrityValidatingStatementOperations.labelGetOrCreateForName(DataIntegrityValidatingStatemen
"org.neo4j.kernel.impl.api.OperationsFacade.labelGetOrCreateForName(OperationsFacade.java:390)",
"org.neo4j.kernel.impl.core.NodeProxy.addLabel(NodeProxy.java:472)",
],
"fullname": "org.neo4j.kernel.api.exceptions.schema.IllegalTokenNameException"
},
"exception": "ConstraintViolationException",
"stacktrace": [
"org.neo4j.kernel.impl.core.NodeProxy.addLabel(NodeProxy.java:481)",
],
"fullname": "org.neo4j.graphdb.ConstraintViolationException"
}
}
19.12.4.Replacing labels on a node

This removes any labels currently on a node, and replaces them with the labels passed in as the
request body.
285
REST API
Node[ 102] : Direct or, Act or
Example request
PUT http://localhost:7474/db/data/node/102/labels
[ "Actor", "Director" ]
Example response
204: No Content
19.12.5.Removing a label from a node

Node[ 103] nam e = 'Clint East wood'
Example request
DELETE http://localhost:7474/db/data/node/103/labels/Person
Example response
204: No Content
19.12.6.Listing labels for a node

Example request
GET http://localhost:7474/db/data/node/108/labels
Example response
200: OK
286
REST API
[ "Director", "Actor" ]
19.12.7.Get all nodes with a label

Node[ 113] : Direct or
nam e = 'St even Spielberg'

Node[ 115] : Act or

nam e = 'Donald Sut herland'
Example request
GET http://localhost:7474/db/data/label/Actor/nodes
Example response
200: OK
[ {
"data" : {
"name" : "Clint Eastwood"
},
"extensions" : {
},
}, {
"data" : {
"name" : "Donald Sutherland"
},
"extensions" : {
},
} ]
287
REST API
19.12.8.Get nodes by label and property

You can retrieve all nodes with a given label and property by passing one property as a query
parameter. Notice that the property value is JSON-encoded and then URL-encoded.
If there is an index available on the label/property combination you send, that index will be used. If no
index is available, all nodes with the given label will be filtered through to find matching nodes.
Currently, it is not possible to search using multiple properties.
Node[ 116] : Person
Node[ 117] : Person
Node[ 118] : Person
Example request
GET http://localhost:7474/db/data/label/Person/nodes?name=%22Clint+Eastwood%22
Example response
200: OK
[ {
"data" : {
"name" : "Clint Eastwood"
},
"extensions" : {
},
} ]
19.12.9.List all labels

Node[ 109] : Person, Direct or
Node[ 110] : Person, Direct or, Act or

Example request
GET http://localhost:7474/db/data/labels
288
Node[ 111] : Person, Act or

nam e = 'Donald Sut herland'
REST API
Example response
200: OK
[ "Actor", "mylabel", "novice", "expert", "person", "SomeLabel", "coder", "Person", "user", "chef", "Director" ]
289
REST API
19.13.Indexing
Note
This documents schema based indexes, a feature that was introduced in Neo4j 2.0, see
Section19.18, Legacy indexing for legacy indexing.
For more details about indexes and the optional schema in Neo4j, see Section3.7, Schema.
19.13.1.Create index
This will start a background job in the database that will create and populate the index. You can check
the status of your index by listing all the indexes for the relevant label. The created index will show
up, but have a state of POPULATING until the index is ready, where it is marked as ONLINE.
Example request
POST http://localhost:7474/db/data/schema/index/person
{
"property_keys" : [ "name" ]
}
Example response
200: OK
{
"label" : "person",
}
19.13.2.List indexes for a label

Example request
GET http://localhost:7474/db/data/schema/index/user
Example response
200: OK
[ {
"property_keys" : [ "name" ],
"label" : "user"
} ]
19.13.3.Drop index
Drop index
Example request
290
REST API
DELETE http://localhost:7474/db/data/schema/index/SomeLabel/name
Example response
204: No Content
291
REST API
19.14.Constraints
19.14.1.Create uniqueness constraint
Create a uniqueness constraint on a property.
Example request
POST http://localhost:7474/db/data/schema/constraint/person/uniqueness/
{
}
Example response
200: OK
{
"label" : "person",
"type" : "UNIQUENESS",
}
19.14.2.Get a specific uniqueness constraint

Get a specific uniqueness constraint for a label and a property.
Example request
GET http://localhost:7474/db/data/schema/constraint/user/uniqueness/name
Example response
200: OK
[ {
"property_keys" : [ "name" ],
"label" : "user",
"type" : "UNIQUENESS"
} ]
19.14.3.Get all uniqueness constraints for a label

Example request
GET http://localhost:7474/db/data/schema/constraint/user/uniqueness/
Example response
200: OK
292
REST API
[ {
"property_keys" : [ "name2" ],
"label" : "user",
}, {
"label" : "user",
} ]
19.14.4.Get all constraints for a label

Example request
GET http://localhost:7474/db/data/schema/constraint/user
Example response
200: OK
[ {
"label" : "user",
}, {
"label" : "user",
} ]
19.14.5.Get all constraints

Example request
GET http://localhost:7474/db/data/schema/constraint/
Example response
200: OK
[ {
"label" : "prog",
}, {
"label" : "user",
} ]
19.14.6.Drop constraint
Drop uniqueness constraint for a label and a property.
293
REST API
Example request
DELETE http://localhost:7474/db/data/schema/constraint/SomeLabel/uniqueness/name
Example response
204: No Content
294
REST API
19.15.Traversals
Warning
The Traversal REST Endpoint executes arbitrary Javascript code under the hood as part of
the evaluators definitions. In hosted and open environments, this can constitute a security
risk. In these case, consider using declarative approaches like PartIII, Cypher Query
Language or write your own server side plugin executing the interesting traversals with
the Java API ( see Section31.1, Server Plugins ) or secure your server, see Chapter25,
Security.
Traversals are performed from a start node. The traversal is controlled by the URI and the body sent
with the request.
returnType
The kind of objects in the response is determined by traverse/{returnType} in the URL. returnType
can have one of these values:
node
relationship
path: contains full representations of start and end node, the rest are URIs.
fullpath: contains full representations of all nodes and relationships.
To decide how the graph should be traversed you can use these parameters in the request body:
order
Decides in which order to visit nodes. Possible values:
breadth_first: see Breadth-first search <http://en.wikipedia.org/wiki/Breadth-first_search>.
depth_first: see Depth-first search <http://en.wikipedia.org/wiki/Depth-first_search>
relationships
Decides which relationship types and directions should be followed. The direction can be one of:
all
in
out
uniqueness
Decides how uniqueness should be calculated. For details on different uniqueness values see the
Java API on Uniqueness <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/kernel/
Uniqueness.html>. Possible values:
node_global
none
relationship_global
node_path
relationship_path
prune_evaluator
Decides whether the traverser should continue down that path or if it should be pruned so that
the traverser wont continue down that path. You can write your own prune evaluator as (see
Section19.15.1, Traversal using a return filter or use the built-in none prune evaluator.
295
REST API
return_filter
Decides whether the current position should be included in the result. You can provide your own
code for this (see Section19.15.1, Traversal using a return filter), or use one of the built-in
filters:
all
all_but_start_node
max_depth
Is a short-hand way of specifying a prune evaluator which prunes after a certain depth. If not
specified a max depth of 1 is used and if a prune_evaluator is specified instead of a max_depth, no
max depth limit is set.
The position object in the body of the return_filter and prune_evaluator is a Path <http://
components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/Path.html> object representing the path
from the start node to the current traversal position.
Out of the box, the REST API supports JavaScript code in filters and evaluators. The script
body will be executed in a Java context which has access to the full Neo4j Java API <http://
components.neo4j.org/neo4j/2.0.0/apidocs/>. See the examples for the exact syntax of the request.
19.15.1.Traversal using a return filter

In this example, the none prune evaluator is used and a return filter is supplied in order to return all
names containing "t". The result is to be returned as nodes and the max depth is set to 3.
296
REST API
Node[ 375] nam e = 'Root '
knows
Node[ 373] nam e = 'Johan'
knows
Node[ 378] nam e = 'Mat t ias'
knows
Node[ 374] nam e = 'Em il'
knows
knows
Node[ 377] nam e = 'Pet er'
Node[ 376] nam e = 'Tobias'
loves
Example request
POST http://localhost:7474/db/data/node/375/traverse/node
{
"order" : "breadth_first",
"return_filter" : {
"body" : "position.endNode().getProperty('name').toLowerCase().contains('t')",
"language" : "javascript"
},
"prune_evaluator" : {
"body" : "position.length() > 10",
},
"uniqueness" : "node_global",
"relationships" : [ {
"direction" : "all",
"type" : "knows"
}, {
"direction" : "all",
"type" : "loves"
} ],
"max_depth" : 3
}
297
REST API
Example response
200: OK
[ {
"data" : {
"name" : "Root"
},
"extensions" : {
},
}, {
"data" : {
"name" : "Mattias"
},
"extensions" : {
},
}, {
"data" : {
"name" : "Peter"
},
"extensions" : {
},
}, {
298
REST API
"data" : {
"name" : "Tobias"
},
"extensions" : {
},
} ]
19.15.2.Return relationships from a traversal

own
know
Node[ 367] nam e = 'car'
Example request
POST http://localhost:7474/db/data/node/369/traverse/relationship
{
"uniqueness" : "none",
"return_filter" : {
"language" : "builtin",
"name" : "all"
}
}
Example response
200: OK
[ {
"data" : {
},
299
REST API
"type" : "know",
"extensions" : {
},
}, {
"data" : {
},
"type" : "own",
"extensions" : {
},
} ]
19.15.3.Return paths from a traversal

own
know
Node[ 357] nam e = 'car'
Example request
POST http://localhost:7474/db/data/node/359/traverse/path
{
"uniqueness" : "none",
"return_filter" : {
"name" : "all"
}
}
Example response
200: OK
[ {
"nodes" : [ "http://localhost:7474/db/data/node/359" ],
"length" : 0,
"relationships" : [ ],
300
REST API
}, {
"length" : 1,
}, {
"length" : 1,
} ]
19.15.4.Traversal returning nodes below a certain depth

Here, all nodes at a traversal depth below 3 are returned.
Node[ 363] nam e = 'Root '
knows
Node[ 361] nam e = 'Johan'
knows
Node[ 366] nam e = 'Mat t ias'
knows
Node[ 362] nam e = 'Em il'
knows
knows
Node[ 364] nam e = 'Tobias'
loves
Example request
POST http://localhost:7474/db/data/node/363/traverse/node
{
301
REST API
"return_filter" : {
"body" : "position.length()<3;",
},
"name" : "none",
"language" : "builtin"
}
}
Example response
200: OK
[ {
"data" : {
"name" : "Root"
},
"extensions" : {
},
}, {
"data" : {
"name" : "Mattias"
},
"extensions" : {
},
}, {
"data" : {
"name" : "Johan"
},
302
REST API
"extensions" : {
},
}, {
"data" : {
"name" : "Emil"
},
"extensions" : {
},
} ]
19.15.5.Creating a paged traverser

Paged traversers are created by POST-ing a traversal description to the link identified by the
paged_traverser key in a node representation. When creating a paged traverser, the same options apply
as for a regular traverser, meaning that node, path, or fullpath, can be targeted.
Example request
POST http://localhost:7474/db/data/node/297/paged/traverse/node
{
"name" : "none"
},
"return_filter" : {
"language" : "javascript",
"body" : "position.endNode().getProperty('name').contains('1');"
},
"order" : "depth_first",
"relationships" : {
"type" : "NEXT",
"direction" : "out"
}
}
Example response
201: Created
303
REST API
Location: http://localhost:7474/db/data/node/297/paged/traverse/node/
ab3224c783f24e1fb77789e0c89ba641
[ {
"data" : {
"name" : "1"
},
"extensions" : {
},
}, {
"data" : {
"name" : "10"
},
"extensions" : {
},
}, {
"data" : {
"name" : "11"
},
"extensions" : {
},
}, {
"data" : {
304
REST API
"name" : "12"
},
"extensions" : {
},
}, {
"data" : {
"name" : "13"
},
"extensions" : {
},
}, {
"data" : {
"name" : "14"
},
"extensions" : {
},
}, {
"data" : {
"name" : "15"
},
305
REST API
"extensions" : {
},
}, {
"data" : {
"name" : "16"
},
"extensions" : {
},
}, {
"data" : {
"name" : "17"
},
"extensions" : {
},
}, {
"data" : {
"name" : "18"
},
"extensions" : {
},
}, {
306
REST API
"data" : {
"name" : "19"
},
"extensions" : {
},
}, {
"data" : {
"name" : "21"
},
"extensions" : {
},
}, {
"data" : {
"name" : "31"
},
"extensions" : {
},
} ]
19.15.6.Paging through the results of a paged traverser

Paged traversers hold state on the server, and allow clients to page through the results of a traversal.
To progress to the next page of traversal results, the client issues a HTTP GET request on the paged
traversal URI which causes the traversal to fill the next page (or partially fill it if insufficient results
are available).
307
REST API
Note that if a traverser expires through inactivity it will cause a 404 response on the next GET request.
Traversers' leases are renewed on every successful access for the same amount of time as originally
specified.
When the paged traverser reaches the end of its results, the client can expect a 404 response as the
traverser is disposed by the server.
Example request
GET http://localhost:7474/db/data/node/330/paged/traverse/node/7ab37c0064654259ab2a1212ddb41c7f
Example response
200: OK
[ {
"data" : {
"name" : "331"
},
"extensions" : {
},
}, {
"data" : {
"name" : "341"
},
"extensions" : {
},
}, {
"data" : {
"name" : "351"
},
308
REST API
"extensions" : {
},
}, {
"data" : {
"name" : "361"
},
"extensions" : {
},
}, {
"data" : {
"name" : "371"
},
"extensions" : {
},
}, {
"data" : {
"name" : "381"
},
"extensions" : {
},
309
REST API
}, {
"data" : {
"name" : "391"
},
"extensions" : {
},
}, {
"data" : {
"name" : "401"
},
"extensions" : {
},
}, {
"data" : {
"name" : "410"
},
"extensions" : {
},
}, {
"data" : {
310
REST API
"name" : "411"
},
"extensions" : {
},
}, {
"data" : {
"name" : "412"
},
"extensions" : {
},
}, {
"data" : {
"name" : "413"
},
"extensions" : {
},
}, {
"data" : {
"name" : "414"
},
311
REST API
"extensions" : {
},
}, {
"data" : {
"name" : "415"
},
"extensions" : {
},
}, {
"data" : {
"name" : "416"
},
"extensions" : {
},
}, {
"data" : {
"name" : "417"
},
"extensions" : {
},
}, {
312
REST API
"data" : {
"name" : "418"
},
"extensions" : {
},
}, {
"data" : {
"name" : "419"
},
"extensions" : {
},
}, {
"data" : {
"name" : "421"
},
"extensions" : {
},
}, {
"data" : {
"name" : "431"
},
313
REST API
"extensions" : {
},
}, {
"data" : {
"name" : "441"
},
"extensions" : {
},
} ]
19.15.7.Paged traverser page size

The default page size is 50 items, but depending on the application larger or smaller pages sizes might
be appropriate. This can be set by adding a pageSize query parameter.
Example request
POST http://localhost:7474/db/data/node/33/paged/traverse/node?pageSize=1
{
"name" : "none"
},
"return_filter" : {
},
"relationships" : {
"type" : "NEXT",
"direction" : "out"
}
}
Example response
201: Created
314
REST API
cccb89c7d1f04ca5a021b5ecbaf2075b
[ {
"data" : {
"name" : "1"
},
"extensions" : {
},
} ]
19.15.8.Paged traverser timeout
The default timeout for a paged traverser is 60 seconds, but depending on the application larger or
smaller timeouts might be appropriate. This can be set by adding a leaseTime query parameter with the
number of seconds the paged traverser should last.
Example request
POST http://localhost:7474/db/data/node/807/paged/traverse/node?leaseTime=10
{
"name" : "none"
},
"return_filter" : {
},
"relationships" : {
"type" : "NEXT",
"direction" : "out"
}
}
Example response
201: Created
e05ab52b5e1b4e7ca3d794cf06fadce2
[ {
315
REST API
"data" : {
"name" : "1"
},
"extensions" : {
},
}, {
"data" : {
"name" : "10"
},
"extensions" : {
},
}, {
"data" : {
"name" : "11"
},
"extensions" : {
},
}, {
"data" : {
"name" : "12"
},
316
REST API
"extensions" : {
},
}, {
"data" : {
"name" : "13"
},
"extensions" : {
},
}, {
"data" : {
"name" : "14"
},
"extensions" : {
},
}, {
"data" : {
"name" : "15"
},
"extensions" : {
},
317
REST API
}, {
"data" : {
"name" : "16"
},
"extensions" : {
},
}, {
"data" : {
"name" : "17"
},
"extensions" : {
},
}, {
"data" : {
"name" : "18"
},
"extensions" : {
},
}, {
"data" : {
"name" : "19"
},
318
REST API
"extensions" : {
},
}, {
"data" : {
"name" : "21"
},
"extensions" : {
},
}, {
"data" : {
"name" : "31"
},
"extensions" : {
},
} ]
319
REST API
19.16.Graph Algorithms
Neo4j comes with a number of built-in graph algorithms. They are performed from a start node. The
traversal is controlled by the URI and the body sent with the request. These are the parameters that
can be used:
algorithm
The algorithm to choose. If not set, default is shortestPath. algorithm can have one of these values:
shortestPath
allSimplePaths
allPaths
dijkstra (optionally with cost_property and default_cost parameters)
max_depth
The maximum depth as an integer for the algorithms like shortestPath, where applicable. Default
is 1.
19.16.1.Find all shortest paths

The shortestPath algorithm can find multiple paths between the same nodes, like in this example.
Node[ 202] nam e = 'a'
to
to
Node[ 201] nam e = 'c'
to
Node[ 198] nam e = 'd'
Node[ 200] nam e = 'b'
to
to
Node[ 199] nam e = 'e'
to
to
Node[ 196] nam e = 'f'
to
to
Node[ 197] nam e = 'g'
Example request
320
to
REST API
POST http://localhost:7474/db/data/node/202/paths
{
"max_depth" : 3,
"relationships" : {
"type" : "to",
"direction" : "out"
},
"algorithm" : "shortestPath"
}
Example response
200: OK
[ {
"nodes" : [ "http://localhost:7474/db/data/node/202", "http://localhost:7474/db/data/node/201", "http://localhost:7474/db/data/node
"length" : 2,
"relationships" : [ "http://localhost:7474/db/data/relationship/138", "http://localhost:7474/db/data/relationship/147" ],
}, {
"length" : 2,
} ]
19.16.2.Find one of the shortest paths

If no path algorithm is specified, a shortestPath algorithm with a max depth of 1 will be chosen. In
this example, the max_depth is set to 3 in order to find the shortest path between a maximum of 3 linked
nodes.
321
REST API
to
to
to
to
to
to
to
to
to
to
Node[ 190] nam e = 'g'
Example request
POST http://localhost:7474/db/data/node/195/path
{
"max_depth" : 3,
"relationships" : {
"type" : "to",
"direction" : "out"
},
"algorithm" : "shortestPath"
}
Example response
200: OK
{
"length" : 2,
322
REST API
}
19.16.3.Execute a Dijkstra algorithm and get a single path

This example is running a Dijkstra algorithm over a graph with different cost properties on different
relationships. Note that the request URI ends with /path which means a single path is what we want
here.
to
cost = 0.5
to
cost = 1.5
to
cost = 0.5
to
cost = 0.5
to
cost = 1.2
to
cost = 0.5
to
cost = 0.5
Example request
{
"cost_property" : "cost",
"relationships" : {
"type" : "to",
"direction" : "out"
},
"algorithm" : "dijkstra"
}
Example response
200: OK
323
REST API
{
"weight" : 1.5,
"length" : 3,
"relationships" : [ "http://localhost:7474/db/data/relationship/149", "http://localhost:7474/db/data/relationship/151", "http://loc
}
19.16.4.Execute a Dijkstra algorithm with equal weights on relationships

The following is executing a Dijkstra search on a graph with equal weights on all relationships. This
example is included to show the difference when the same graph structure is used, but the path weight
is equal to the number of hops.
to
cost = 1
to
cost = 1
to
cost = 1
to
cost = 1
to
cost = 1
to
cost = 1
to
cost = 1
Example request
{
"relationships" : {
"type" : "to",
324
REST API
"direction" : "out"
},
}
Example response
200: OK
{
"weight" : 2.0,
"length" : 2,
}
19.16.5.Execute a Dijkstra algorithm and get multiple paths

This example is running a Dijkstra algorithm over a graph with different cost properties on different
relationships. Note that the request URI ends with /paths which means we want multiple paths
returned, in case they exist.
to
cost = 0.5
to
cost = 1.5
to
cost = 0.5
to
cost = 0.5
to
cost = 1.0
to
cost = 0.5
to
cost = 0.5
Example request
POST http://localhost:7474/db/data/node/188/paths
325
REST API
{
"relationships" : {
"type" : "to",
"direction" : "out"
},
}
Example response
200: OK
[ {
"weight" : 1.5,
"length" : 3,
"relationships" : [ "http://localhost:7474/db/data/relationship/122", "http://localhost:7474/db/data/relationship/124", "http://loc
}, {
"weight" : 1.5,
"length" : 2,
} ]
326
REST API
19.17.Batch operations
19.17.1.Execute multiple operations in batch
This lets you execute multiple API calls through a single HTTP call, significantly improving
performance for large insert and update operations.
The batch service expects an array of job descriptions as input, each job description describing an
action to be performed via the normal server API.
This service is transactional. If any of the operations performed fails (returns a non-2xx HTTP status
code), the transaction will be rolled back and all changes will be undone.
Each job description should contain a to attribute, with a value relative to the data API root (so http://
localhost:7474/db/data/node becomes just /node), and a method attribute containing HTTP verb to use.
Optionally you may provide a body attribute, and an id attribute to help you keep track of responses,
although responses are guaranteed to be returned in the same order the job descriptions are received.
The following figure outlines the different parts of the job descriptions:
Node[ 31] age = 1
Node[ 32] age = 1
knows
Example request
POST http://localhost:7474/db/data/batch
[ {
"method" : "PUT",
"to" : "/node/31/properties",
"body" : {
"age" : 1
},
"id" : 0
}, {
"method" : "GET",
327
Node[ 33] age = 1
REST API
"to" : "/node/31",
"id" : 1
}, {
"method" : "POST",
"to" : "/node",
"body" : {
"age" : 1
},
"id" : 2
}, {
"method" : "POST",
"to" : "/node",
"body" : {
"age" : 1
},
"id" : 3
} ]
Example response
200: OK
[ {
"id" : 0,
"from" : "/node/31/properties"
}, {
"id" : 1,
"body" : {
"extensions" : {
},
"data" : {
"age" : 1
}
},
"from" : "/node/31"
}, {
"id" : 2,
"location" : "http://localhost:7474/db/data/node/32",
"body" : {
"extensions" : {
},
328
REST API
"data" : {
"age" : 1
}
},
"from" : "/node"
}, {
"id" : 3,
"body" : {
"extensions" : {
},
"data" : {
"age" : 1
}
},
"from" : "/node"
} ]
19.17.2.Refer to items created earlier in the same batch job

The batch operation API allows you to refer to the URI returned from a created resource in subsequent
job descriptions, within the same batch call.
Use the {[JOB ID]} special syntax to inject URIs from created resources into JSON strings in
subsequent job descriptions.
Node[ 23] nam e = 'bob'
KNOWS
since = '2010'
Node[ 24] age = 12
Example request
329
REST API
[ {
"method" : "POST",
"to" : "/node",
"id" : 0,
"body" : {
"name" : "bob"
}
}, {
"method" : "POST",
"to" : "/node",
"id" : 1,
"body" : {
"age" : 12
}
}, {
"method" : "POST",
"to" : "{0}/relationships",
"id" : 3,
"body" : {
"to" : "{1}",
"data" : {
"since" : "2010"
},
"type" : "KNOWS"
}
}, {
"method" : "POST",
"to" : "/index/relationship/my_rels",
"id" : 4,
"body" : {
"key" : "since",
"value" : "2010",
"uri" : "{3}"
}
} ]
Example response
200: OK
[ {
"id" : 0,
"body" : {
"extensions" : {
},
"data" : {
"name" : "bob"
}
},
330
REST API
"from" : "/node"
}, {
"id" : 1,
"body" : {
"extensions" : {
},
"data" : {
"age" : 12
}
},
"from" : "/node"
}, {
"id" : 3,
"location" : "http://localhost:7474/db/data/relationship/3",
"body" : {
"extensions" : {
},
"type" : "KNOWS",
"data" : {
"since" : "2010"
}
},
"from" : "http://localhost:7474/db/data/node/23/relationships"
}, {
"id" : 4,
"location" : "http://localhost:7474/db/data/index/relationship/my_rels/since/2010/3",
"body" : {
"extensions" : {
},
"type" : "KNOWS",
"data" : {
"since" : "2010"
},
"indexed" : "http://localhost:7474/db/data/index/relationship/my_rels/since/2010/3"
},
"from" : "/index/relationship/my_rels"
} ]
331
REST API
19.17.3.Execute multiple operations in batch streaming

Node[ 101] age = 1
Node[ 102] age = 1
knows
Example request
X-Stream: true
[ {
"method" : "PUT",
"to" : "/node/101/properties",
"body" : {
"age" : 1
},
"id" : 0
}, {
"method" : "GET",
"to" : "/node/101",
"id" : 1
}, {
"method" : "POST",
"to" : "/node",
"body" : {
"age" : 1
},
"id" : 2
}, {
"method" : "POST",
"to" : "/node",
"body" : {
"age" : 1
},
"id" : 3
} ]
Example response
200: OK
[ {
"id" : 0,
"from" : "/node/101/properties",
"body" : null,
"status" : 204
}, {
"id" : 1,
332
Node[ 103] age = 1
REST API
"from" : "/node/101",
"body" : {
"extensions" : {
},
"data" : {
"age" : 1
}
},
"status" : 200
}, {
"id" : 2,
"from" : "/node",
"body" : {
"extensions" : {
},
"data" : {
"age" : 1
}
},
"status" : 201
}, {
"id" : 3,
"from" : "/node",
"body" : {
"extensions" : {
},
333
REST API
"data" : {
"age" : 1
}
},
"status" : 201
} ]
334
REST API
19.18.Legacy indexing
Note
This documents the legacy indexing in Neo4j, which is no longer the preferred way to
handle indexes. Consider looking at Section19.13, Indexing.
An index can contain either nodes or relationships.
Note
To create an index with default configuration, simply start using it by adding nodes/
relationships to it. It will then be automatically created for you.
What default configuration means depends on how you have configured your database. If you havent
changed any indexing configuration, it means the indexes will be using a Lucene-based backend.
All the examples below show you how to do operations on node indexes, but all of them are just as
applicable to relationship indexes. Simple change the "node" part of the URL to "relationship".
If you want to customize the index settings, see Section19.18.2, Create node index with
configuration.
19.18.1.Create node index

Note
Instead of creating the index this way, you can simply start to use it, and it will be created
automatically with default configuration.
Example request
POST http://localhost:7474/db/data/index/node/
{
"name" : "favorites"
}
Example response
201: Created
Location: http://localhost:7474/db/data/index/node/favorites/
{
"template" : "http://localhost:7474/db/data/index/node/favorites/{key}/{value}"
}
19.18.2.Create node index with configuration

This request is only necessary if you want to customize the index settings. If you are happy with the
defaults, you can just start indexing nodes/relationships, as non-existent indexes will automatically be
created as you do. See Section34.10, Configuration and fulltext indexes for more information on
index configuration.
Example request
335
REST API
{
"name" : "fulltext",
"config" : {
"type" : "fulltext",
"provider" : "lucene"
}
}
Example response
201: Created
Location: http://localhost:7474/db/data/index/node/fulltext/
{
"template" : "http://localhost:7474/db/data/index/node/fulltext/{key}/{value}",
}
19.18.3.Delete node index

Example request
DELETE http://localhost:7474/db/data/index/node/kvnode
Example response
204: No Content
19.18.4.List node indexes

Example request
GET http://localhost:7474/db/data/index/node/
Example response
200: OK
{
"node_auto_index" : {
"template" : "http://localhost:7474/db/data/index/node/node_auto_index/{key}/{value}",
"provider" : "lucene",
"type" : "exact"
},
"favorites" : {
"template" : "http://localhost:7474/db/data/index/node/favorites/{key}/{value}",
"provider" : "lucene",
336
REST API
"type" : "exact"
}
}
19.18.5.Add node to index

Associates a node with the given key/value pair in the given index.
Note
Spaces in the URI have to be encoded as %20.
Caution
This does not overwrite previous entries. If you index the same key/value/item
combination twice, two index entries are created. To do update-type operations, you need
to delete the old entry before adding a new one.
Example request
POST http://localhost:7474/db/data/index/node/favorites
{
"value" : "some value",
"uri" : "http://localhost:7474/db/data/node/83",
"key" : "some-key"
}
Example response
201: Created
Location: http://localhost:7474/db/data/index/node/favorites/some-key/some%20value/83
{
"extensions" : {
},
"data" : {
},
"indexed" : "http://localhost:7474/db/data/index/node/favorites/some-key/some%20value/83"
}
19.18.6.Remove all entries with a given node from an index

Example request
337
REST API
DELETE http://localhost:7474/db/data/index/node/kvnode/89
Example response
204: No Content
19.18.7.Remove all entries with a given node and key from an index
Example request
DELETE http://localhost:7474/db/data/index/node/kvnode/kvkey2/91
Example response
204: No Content
19.18.8.Remove all entries with a given node, key and value from an index
Example request
DELETE http://localhost:7474/db/data/index/node/kvnode/kvkey1/value1/84
Example response
204: No Content
19.18.9.Find node by exact match

Note
Spaces in the URI have to be encoded as %20.

Example request
GET http://localhost:7474/db/data/index/node/favorites/key/the%2520value
Example response
200: OK
[ {
"indexed" : "http://localhost:7474/db/data/index/node/favorites/key/the%2520value/100",
"data" : {
},
338
REST API
"extensions" : {
},
} ]
19.18.10.Find node by query

The query language used here depends on what type of index you are querying. The default index
type is Lucene, in which case you should use the Lucene query language here. Below an example of a
fuzzy search over multiple keys.
See: http://lucene.apache.org/core/3_6_2/queryparsersyntax.html
Getting the results with a predefined ordering requires adding the parameter
order=ordering
where ordering is one of index, relevance or score. In this case an additional field will be added to
each result, named score, that holds the float value that is the score reported by the query result.
Example request
GET http://localhost:7474/db/data/index/node/bobTheIndex?query=Name:Build~0.1%20AND
%20Gender:Male

Example response
200: OK
[ {
"data" : {
"Name" : "Builder"
},
"extensions" : {
},
} ]
339
REST API
19.19.Unique Indexing
Note
As of Neo4j 2.0, unique constraints have been added. These make Neo4j enforce the
uniqueness, guaranteeing that uniqueness is maintained. See Section3.7.2, Constraints
for details about this. For most cases, the unique constraints should be used rather than the
features described below.
For uniqueness enforcements, there are two modes:
URL Parameter uniqueness=get_or_create: Create a new node/relationship and index it if no existing
one can be found. If an existing node/relationship is found, discard the sent data and return the
existing node/relationship.
URL Parameter uniqueness=create_or_fail: Create a new node/relationship if no existing one can be
found in the index. If an existing node/relationship is found, return a conflict error.
For more information, see Section16.6, Creating unique nodes.
19.19.1.Get or create unique node (create)

The node is created if it doesnt exist in the unique index already.
Example request
POST http://localhost:7474/db/data/index/node/people?uniqueness=get_or_create
{
"key" : "name",
"value" : "Tobias",
"properties" : {
"name" : "Tobias",
"sequence" : 1
}
}
Example response
201: Created
Location: http://localhost:7474/db/data/index/node/people/name/Tobias/98
{
"extensions" : {
},
340
REST API
"data" : {
"sequence" : 1,
"name" : "Tobias"
},
"indexed" : "http://localhost:7474/db/data/index/node/people/name/Tobias/98"
}
19.19.2.Get or create unique node (existing)

Here, a node is not created but the existing unique node returned, since another node is indexed with
the same data already. The node data returned is then that of the already existing node.
Example request
POST http://localhost:7474/db/data/index/node/people?uniqueness=get_or_create
{
"key" : "name",
"value" : "Peter",
"properties" : {
"name" : "Peter",
"sequence" : 2
}
}
Example response
200: OK
Location: http://localhost:7474/db/data/index/node/people/name/Peter/88
{
"extensions" : {
},
"data" : {
"sequence" : 1,
"name" : "Peter"
},
"indexed" : "http://localhost:7474/db/data/index/node/people/name/Peter/88"
}
19.19.3.Create a unique node or return fail (create)

Here, in case of an already existing node, an error should be returned. In this example, no existing
indexed node is found and a new node is created.
341
REST API
Example request
POST http://localhost:7474/db/data/index/node/people?uniqueness=create_or_fail
{
"key" : "name",
"value" : "Tobias",
"properties" : {
"name" : "Tobias",
"sequence" : 1
}
}
Example response
201: Created
Location: http://localhost:7474/db/data/index/node/people/name/Tobias/97
{
"extensions" : {
},
"data" : {
"sequence" : 1,
"name" : "Tobias"
},
"indexed" : "http://localhost:7474/db/data/index/node/people/name/Tobias/97"
}
19.19.4.Create a unique node or return fail (fail)

Here, in case of an already existing node, an error should be returned. In this example, an existing
node indexed with the same data is found and an error is returned.
Example request
POST http://localhost:7474/db/data/index/node/people?uniqueness=create_or_fail
{
"key" : "name",
"value" : "Peter",
"properties" : {
342
REST API
"name" : "Peter",
"sequence" : 2
}
}
Example response
409: Conflict
{
"extensions" : {
},
"data" : {
"sequence" : 1,
"name" : "Peter"
},
"indexed" : "http://localhost:7474/db/data/index/node/people/name/Peter/81"
}
19.19.5.Add an existing node to unique index (not indexed)

Associates a node with the given key/value pair in the given unique index.
In this example, we are using create_or_fail uniqueness.
Example request
POST http://localhost:7474/db/data/index/node/favorites?uniqueness=create_or_fail
{
"key" : "some-key"
}
Example response
201: Created
Location: http://localhost:7474/db/data/index/node/favorites/some-key/some%20value/93
{
"extensions" : {
},
343
REST API
"data" : {
"some-key" : "some value"
},
}
19.19.6.Add an existing node to unique index (already indexed)

In this case, the node already exists in the index, and thus we get a HTTP 409 status response, as we
have set the uniqueness to create_or_fail.
Example request
POST http://localhost:7474/db/data/index/node/favorites?uniqueness=create_or_fail
{
"key" : "some-key"
}
Example response
409: Conflict
{
"extensions" : {
},
"data" : {
"some-key" : "some value"
},
}
344
REST API
19.19.7.Get or create unique relationship (create)

Create a unique relationship in an index. If a relationship matching the given key and value already
exists in the index, it will be returned. If not, a new relationship will be created.
Note
The type and direction of the relationship is not regarded when determining uniqueness.
Example request
POST http://localhost:7474/db/data/index/relationship/MyIndex/?uniqueness=get_or_create
{
"key" : "name",
"value" : "Tobias",
"type" : "knowledge"
}
Example response
201: Created
Location: http://localhost:7474/db/data/index/relationship/MyIndex/name/Tobias/170
{
"extensions" : {
},
"type" : "knowledge",
"data" : {
"name" : "Tobias"
},
"indexed" : "http://localhost:7474/db/data/index/relationship/MyIndex/name/Tobias/170"
}
19.19.8.Get or create unique relationship (existing)

Here, in case of an already existing relationship, the sent data is ignored and the existing relationship
returned.
Example request
POST http://localhost:7474/db/data/index/relationship/rels?uniqueness=get_or_create
{
"key" : "name",
"value" : "Peter",
345
REST API
"type" : "KNOWS"
}
Example response
200: OK
{
"extensions" : {
},
"type" : "KNOWS",
"data" : {
},
"indexed" : "http://localhost:7474/db/data/index/relationship/rels/name/Peter/171"
}
19.19.9.Create a unique relationship or return fail (create)

Here, in case of an already existing relationship, an error should be returned. In this example, no
existing relationship is found and a new relationship is created.
Example request
POST http://localhost:7474/db/data/index/relationship/rels?uniqueness=create_or_fail
{
"key" : "name",
"value" : "Tobias",
"type" : "KNOWS"
}
Example response
201: Created
Location: http://localhost:7474/db/data/index/relationship/rels/name/Tobias/174
{
"extensions" : {
},
"type" : "KNOWS",
"data" : {
"name" : "Tobias"
},
346
REST API
"indexed" : "http://localhost:7474/db/data/index/relationship/rels/name/Tobias/174"
}
19.19.10.Create a unique relationship or return fail (fail)

Here, in case of an already existing relationship, an error should be returned. In this example, an
existing relationship is found and an error is returned.
Example request
{
"key" : "name",
"value" : "Peter",
"type" : "KNOWS"
}
Example response
409: Conflict
{
"extensions" : {
},
"type" : "KNOWS",
"data" : {
},
}
19.19.11.Add an existing relationship to a unique index (not indexed)

If a relationship matching the given key and value already exists in the index, it will be returned. If
not, an HTTP 409 (conflict) status will be returned in this case, as we are using create_or_fail.
Its possible to use get_or_create uniqueness as well.
Note
The type and direction of the relationship is not regarded when determining uniqueness.
Example request
{
347
REST API
"key" : "name",
"value" : "Peter",
"uri" : "http://localhost:7474/db/data/relationship/164"
}
Example response
201: Created
Location: http://localhost:7474/db/data/index/relationship/rels/name/Peter/165
{
"extensions" : {
},
"type" : "KNOWS",
"data" : {
"name" : "Peter"
},
}
19.19.12.Add an existing relationship to a unique index (already indexed)

Example request
{
"key" : "name",
"value" : "Peter",
"uri" : "http://localhost:7474/db/data/relationship/168"
}
Example response
409: Conflict
{
"extensions" : {
},
"type" : "KNOWS",
"data" : {
},
}
348
REST API
19.20.Legacy Automatic Indexes

To enable automatic indexes, set up the database for that, see Section34.12.1, Configuration. With
this feature enabled, you can then index and query nodes in these indexes.
19.20.1.Find node by exact match from an automatic index

Automatic index nodes can be found via exact lookups with normal Index REST syntax.
Example request
GET http://localhost:7474/db/data/index/auto/node/name/I
Example response
200: OK
[ {
"data" : {
"name" : "I"
},
"extensions" : {
},
} ]
19.20.2.Find node by query from an automatic index

See Find node by query for the actual query syntax.
Example request
GET http://localhost:7474/db/data/index/auto/node/?query=name:I
Example response
200: OK
[ {
"data" : {
"name" : "I"
349
REST API
},
"extensions" : {
},
} ]
350
REST API
19.21.Configurable Legacy Automatic Indexing

Out of the box auto-indexing supports exact matches since they are created with the default
configuration (see Section34.12, Automatic Indexing) the first time you access them. However it is
possible to intervene in the lifecycle of the server before any auto indexes are created to change their
configuration.
Warning
This approach cannot be used on databases that already have auto-indexes established. To
change the auto-index configuration existing indexes would have to be deleted first, so be
careful!
Caution
This technique works, but it is not particularly pleasant. Future versions of Neo4j may
remove this loophole in favour of a better structured feature for managing auto-indexing
configurations.
Auto-indexing must be enabled through configuration before we can create or configure them. Firstly
ensure that youve added some config like this into your servers neo4j.properties file:
node_auto_indexing=true
relationship_auto_indexing=true
node_keys_indexable=name,phone
relationship_keys_indexable=since
The node_auto_indexing and relationship_auto_indexing settings turn auto-indexing on for nodes and
relationships respectively. The node_keys_indexable key allows you to specify a comma-separated list
of node property keys to be indexed. The relationship_keys_indexable does the same for relationship
property keys.
Next start the server as usual by invoking the start script as described in Section21.2, Server
Installation.
Next we have to pre-empt the creation of an auto-index, by telling the server to create an apparently
manual index which has the same name as the node (or relationship) auto-index. For example, in this
case well create a node auto index whose name is node_auto_index, like so:
19.21.1.Create an auto index for nodes with specific configuration

Example request
{
"name" : "node_auto_index",
"config" : {
}
}
Example response
201: Created
351
REST API
Location: http://localhost:7474/db/data/index/node/node_auto_index/
{
"template" : "http://localhost:7474/db/data/index/node/node_auto_index/{key}/{value}",
}
If you require configured auto-indexes for relationships, the approach is similar:
19.21.2.Create an auto index for relationships with specific configuration

Example request
POST http://localhost:7474/db/data/index/relationship/
{
"name" : "relationship_auto_index",
"config" : {
}
}
Example response
201: Created
Location: http://localhost:7474/db/data/index/relationship/relationship_auto_index/
{
"template" : "http://localhost:7474/db/data/index/relationship/relationship_auto_index/{key}/{value}",
}
In case youre curious how this works, on the server side it triggers the creation of an index which
happens to have the same name as the auto index that the database would create for itself. Now when
we interact with the database, the index thinks the index is already created so the state machine skips
over that step and just gets on with normal day-to-day auto-indexing.
Caution
You have to do this early in your server lifecycle, before any normal auto indexes are
created.
There are a few REST calls providing a REST interface to the AutoIndexer <http://
components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/index/AutoIndexer.html> component.
The following REST calls work both, for node and relationship by simply changing the respective part
of the URL.
19.21.3.Get current status for autoindexing on nodes

Example request
352
REST API
GET http://localhost:7474/db/data/index/auto/node/status
Example response
200: OK
false
19.21.4.Enable node autoindexing

Example request
PUT http://localhost:7474/db/data/index/auto/node/status
true
Example response
204: No Content
19.21.5.Lookup list of properties being autoindexed

Example request
GET http://localhost:7474/db/data/index/auto/node/properties
Example response
200: OK
[ "some-property" ]
19.21.6.Add a property for autoindexing on nodes

Example request
POST http://localhost:7474/db/data/index/auto/node/properties
myProperty1
Example response
204: No Content
19.21.7.Remove a property for autoindexing on nodes

Example request
353
REST API
DELETE http://localhost:7474/db/data/index/auto/node/properties/myProperty1
Example response
204: No Content
354
REST API
19.22.WADL Support
The Neo4j REST API is a truly RESTful interface relying on hypermedia controls (links) to advertise
permissible actions to users. Hypermedia is a dynamic interface style where declarative constructs
(semantic markup) are used to inform clients of their next legal choices just in time.
Caution
RESTful APIs cannot be modeled by static interface description languages like WSDL or
WADL.
However for some use cases, developers may wish to expose WADL descriptions of the Neo4j REST
API, particularly when using tooling that expects such.
In those cases WADL generation may be enabled by adding to your servers neo4j.properties file:
unsupported_wadl_generation_enabled=true
Caution
WADL is not an officially supported part of the Neo4j server API because WADL is
insufficiently expressive to capture the set of potential interactions a client can drive
with Neo4j server. Expect the WADL description to be incomplete, and in some cases
contradictory to the real API. In any cases where the WADL description disagrees with the
REST API, the REST API should be considered authoritative. WADL generation may be
withdrawn at any point in the Neo4j release cycle.
355
Chapter20.Deprecations
This section outlines deprecations in Neo4j 2.0 or earlier in order to help you find a smoother
transition path to future releases.
Graph Matching
The graph-matching component will be removed in future releases.
356
PartV.Operations
This part describes how to install and maintain a Neo4j installation. This includes topics such as backing
up the database and monitoring the health of the database as well as diagnosing issues.
Chapter21.Installation & Deployment

Neo4j is accessed as a standalone server, either directly through a REST interface or through a
language-specific driver.
Neo4j can be installed as a server, running either as a headless application or system service. For
information on installing The Neo4j Server, see Section21.2, Server Installation.
For running Neo4j in high availability mode, see Chapter23, High Availability.
358
Installation & Deployment
21.1.System Requirements
Memory constrains graph size, disk I/O constrains read/write performance, as always.
21.1.1.CPU
Performance is generally memory or I/O bound for large graphs, and compute bound for graphs which
fit in memory.
Minimum
Intel Core i3
Recommended
Intel Core i7
21.1.2.Memory
More memory allows even larger graphs, but runs the risk of inducing larger Garbage Collection
operations.
Minimum
2GB
Recommended
1632GB or more
21.1.3.Disk
Aside from capacity, the performance characteristics of the disk are the most important when selecting
storage.
Minimum
10GB SATA
Recommended
SSD w/ SATA
21.1.4.Filesystem
For proper ACID behavior, the filesystem must support flush (fsync, fdatasync).
Minimum
ext4 (or similar)
Recommended
ext4, ZFS
21.1.5.Software
Neo4j is Java-based.
Java
Oracle Java 7 <http://www.oracle.com/technetwork/java/javase/downloads/index.html>
Operating Systems
Linux, HP UX, Windows 2008 for production; additionally Windows XP, Mac OS X for
development.
359
21.2.Server Installation
21.2.1.Deployment Scenarios
As a developer, you may wish to download Neo4j and run it locally on your desktop computer. We
recommend this as an easy way to discover Neo4j.
For Windows, see Section21.2.3, Windows.
For Unix/Linux, see Section21.2.4, Linux.
For OSX, see Section21.2.5, Mac OSX.
As a systems administrator, you may wish to install Neo4j using a packaging system so you can
ensure that a cluster of machines have identical installs. See the section called Linux Packages for
more information on this.
For information on High Availability, please refer to Chapter23, High Availability.
21.2.2.Prerequisites
With the exception of our Windows Installer, youll need a Java Virtual Machine installed on
your computer. We recommend that you install the Java JDK 7 from the Oracle Website <http://
www.oracle.com/technetwork/java/javase/downloads/index.html>.
21.2.3.Windows
Windows Installer
1. Download the version that you want from http://neo4j.org/download.
Select the appropriate version and architecture for your platform.
2. Double-click the downloaded installer file.
3. Follow the prompts.
Note
The installer will prompt to be granted Administrator privileges. Newer versions

of Windows come with a SmartScreen feature that may prevent the installer from
running you can make it run anyway by clicking "More info" on the "Windows
protected your PC" screen.
Windows Console Application
1. Download the latest release from http://neo4j.org/download.
Select the appropriate Zip distribution.
2. Right-click the downloaded file, click Extract All.
Refer to the top-level extracted directory as: NEO4J_HOME
3. Double-click on %NEO4J_HOME%\bin\Neo4j.bat
4. Stop the server by typing Ctrl-C in the console that got opened.
As a Windows service
Select the appropriate Zip distribution.
360

2. Right-click the downloaded file, click Extract All.
With administrative rights, Neo4j can be installed as a Windows service.
1.
2.
3.
4.
5.
Click Start All Programs Accessories.

Right click Command Prompt Run as Administrator.
Provide authorization and/or the Administrator password.
Navigate to: %NEO4J_HOME%
Run: bin\Neo4jInstaller.bat install
To uninstall, run bin\Neo4jInstaller.bat remove as Administrator.

To query the status of the service, run: bin\Neo4jInstaller.bat status
To start the service from the command prompt, run: sc start Neo4j-Server
To stop the service from the command prompt, run: sc stop Neo4j-Server
You can also start and stop Neo4j from the Windows Service Manager.
Note
Some users have reported problems on Windows when using the ZoneAlarm firewall. If
you are having problems getting large responses from the server, or if the web interface
does not work, try disabling ZoneAlarm. Contact ZoneAlarm support to get information on
how to resolve this.
21.2.4.Linux
Linux Packages
For Debian packages, see the instructions at http://debian.neo4j.org.
Unix Console Application
Select the appropriate tar.gz distribution for your platform.
2. Extract the contents of the archive, using: tar -cf <filename>
3. Change directory to: $NEO4J_HOME
Run: ./bin/neo4j console
4. Stop the server by typing Ctrl-C in the console.
Linux Service
Neo4j can participate in the normal system startup and shutdown process. The following procedure
should work on most popular Linux distributions:
1. cd $NEO4J_HOME
2. sudo ./bin/neo4j-installer install
If asked, enter your password to gain super-user privileges.
3. sudo service neo4j-service status
This should indicate that the server is not running.
361

4. sudo service neo4j-service start
This will start the server.
5. sudo service neo4j-service stop
This will stop the server.
During installation you will be given the option to select the user Neo4j will run as. You will be
asked to supply a username (defaulting to neo4j) and if that user is not present on the system it will be
created as a system account and the $NEO4J_HOME/data directory will be chown'ed to that user.
You are encouraged to create a dedicated user for running the service and for that reason it is
suggested that you unpack the distribution package under /opt or your site specific optional packages
directory.
After installation you may have to do some platform specific configuration and performance tuning.
For that, refer to Section22.12, Linux Performance Guide.
To remove the server from the set of startup services, the proper commands are:
1. cd $NEO4J_HOME
2. sudo ./bin/neo4j-installer remove
This will stop the server, if running, and remove it.
Note that if you chose to create a new user account, on uninstall you will be prompted to remove it
from the system.
Note
This approach to running Neo4j as a server is deprecated. We strongly advise you to run
Neo4j from a package where feasible.
You can alternatively build your own init.d script. See for instance the Linux Standard Base
specification on system initialization <http://refspecs.linuxfoundation.org/LSB_3.1.0/LSB-Coregeneric/LSB-Core-generic/tocsysinit.html>, or one of the many samples <https://gist.github.com/
chrisvest/7673244> and tutorials <http://www.linux.com/learn/tutorials/442412-managing-linuxdaemons-with-init-scripts>.
21.2.5.Mac OSX
OSX via Homebrew
Using Homebrew (see http://brew.sh/), to install the latest stable version of Neo4j Server, issue the
following command:
brew install neo4j && neo4j start
This will get a Neo4j instance running on http://localhost:7474. The installation files will reside in
ls /usr/local/Cellar/neo4j/community-{NEO4J_VERSION}/libexec/ to tweak settings and symlink the
database directory if desired.
After the installation, Neo4j can run either as a service or from a terminal.
Running Neo4j from the Terminal
The server can be started in the background from the terminal with the command neo4j start, and
then stopped again with neo4j stop. The server can also be started in the foreground with neo4j
console then its log output will be printed to the terminal.
362

The neo4j-shell command can be used to interact with Neo4j from the command line using Cypher.
It will automatically connect to any server that is running on localhost with the default port, otherwise
it will show a help message. You can alternatively start the shell with an embedded Neo4j instance,
by using the -path path/to/data argument note that only a single instance of Neo4j can access the
database files at a time.
OSX Service
Neo4j can be installed as a Mac launchd job:
1. cd $NEO4J_HOME
2. ./bin/neo4j-installer install
3. launchctl list | grep neo
This should reveal the launchd "org.neo4j.server.7474" job for running the Neo4j Server.
4. launchctl list | grep neo4j
This should indicate that the server is running.
5. launchctl stop org.neo4j.server
This should stop the server.
6. launchctl start org.neo4j.server
This should start the server again.
To remove the launchctl service, issue the following command:
./bin/neo4j-installer remove
A note on Java on OS X Mavericks

Unlike previous versions, OS X Mavericks does not come with Java pre-installed. You might
encounter that the first time you run Neo4j, where OS X will trigger a popup offering you to install
Java SE 6.
Java SE 6 is incompatible with Neo4j 2.0, so we strongly advise you to skip installing Java SE 6 if
you have no other uses for it. Instead, for Neo4j 2.0 we recommend you install Java SE 7 from Oracle
(http://www.oracle.com/technetwork/java/javase/downloads/index.html) as that is what we support for
production use.
21.2.6.Multiple Server instances on one machine

Neo4j can be set up to run as several instances on one machine, providing for instance several
databases for development.
For how to set this up, see Section23.6.3, Alternative setup: Creating a local cluster for testing. Just
use the Neo4j edition of your choice, follow the guide and remember to not set the servers to run in
HA mode.
363
21.3.Server Installation in the Cloud

Neo4j can be hosted on various cloud services. See http://www.neo4j.org/develop/cloud for more
information.
364
21.4.Upgrading
A database can be upgraded from a minor version to the next, e.g. 1.1 1.2, and 1.2 1.3, but you
can not jump directly from 1.1 1.3. For version 1.8 in particular, it is possible to upgrade directly
from version 1.5.3 and later, as an explicit upgrade. The upgrade process is a one way step; databases
cannot be downgraded.
For most upgrades, only small changes are required to the database store, and these changes proceed
automatically when you start up the database using the newer version of Neo4j.
However, some upgrades require more significant changes to the database store. In these cases, Neo4j
will refuse to start without explicit configuration to allow the upgrade.
The table below lists recent Neo4j versions, and the type of upgrade required.
Upgrade process for Neo4j version
From Version To Version
Upgrade Type
1.3
1.4
Automatic
1.4
1.5
Explicit
1.5
1.6
Explicit
1.6
1.7
Automatic
1.7
1.8
Automatic
1.8
1.9
Automatic
1.9
2.0
Explicit
Note
Downgrade is supported only between versions which do not have incompatible store
layouts. That means that if you did an upgrade where you didnt have to explicitly set the
allow_store_upgrade flag to false then you can downgrade without any problems to the
previous version used. Otherwise downgrading is not supported. In any case, downgrading
currently cannot be done in a rolling fashion, even in HA deployments. Instead, the whole
cluster must be shutdown and each machine downgraded individually and then the service
can be resumed.
21.4.1.Automatic Upgrade
To perform a normal upgrade (for minor changes to the database store):
1.
2.
3.
4.
Download the newer version of Neo4j.

Cleanly shut down the database to upgrade, if it is running.
Startup the database with the newer version of Neo4j.
The upgrade will happen during startup and the process is done when the database has been
successfully started.
21.4.2.Explicit Upgrade
To perform a special upgrade (for significant changes to the database store):
1. Download the newer version of Neo4j.
365

2. Cleanly shut down the database to upgrade, if it is running.
3. Set the Neo4j configuration parameter allow_store_upgrade=true in your neo4j.properties or
embedded configuration.
4. Startup the database with the newer version of Neo4j.
5. The upgrade will happen during startup and the process is done when the database has been
successfully started.
6. The allow_store_upgrade configuration parameter should be removed, set to false or commented
out.
7. Information about the upgrade and progress indicator is printed in messages.log.
21.4.3.Upgrade 1.9 2.0

This edition adds a new store for labels and one for schema, an index for labels, and also the format of
the node store has changed. Note that we do not currently support rolling upgrades between 1.9.x and
2.0.
For Neo4j 2.0, Java 7 is required. We recommend that you install the Java JDK 7 from the Oracle
Website <http://www.oracle.com/technetwork/java/javase/downloads/index.html>.
When upgrading, the following will happen and will both take some time, and will require extra disk
space:
1.
2.
3.
4.
A new, empty schema store will be created.

A new, empty label store will be created.
A new, empty label scan index will be created.
The node store will be converted, we need to enlarge each record to make space for labels this
will take time depending on the size of your database.
Cypher compatibility
Unless you set the statements to still use Cypher 1.9, they have to be updated for the following
changes:
Pattern syntax
Parentheses are required around node patterns when labels are used. This means that when adding
labels to a pattern like a-->b you should use something like (a:Person)-->(b:Company). Its good
practice to use parentheses in node patterns even where they are not strictly required, to enhance
readability.
Optional relationships
The syntax (a)-[?]->(x) for optional relationships has been removed. Use OPTIONAL MATCH instead
(see the corresponding chapter in the Neo4j Manual).
The ! and ? property operators
Expressions like node.property = "value" will not fail when a node without the property is
encountered, and will instead return NULL. This is the same behavior as node.property! = "value"
in Cypher 1.9. The ! property operator has been removed in 2.0. Support for expressions using
the ? property operator, such as node.property? = "value", have also been removed. You can use
not(has(node.property)) OR node.property = "value" instead, which is compatible with both 1.9
and 2.0.
syntax
The CREATE a={foo:bar} syntax has been removed. Instead, use CREATE (a {foo:bar}).
CREATE
366

Using DELETE to remove properties
The DELETE a.prop syntax has been removed. Instead, use REMOVE a.prop.
Using parameters for index keys
Parameters can not be used as the key in START clauses using indexes (for example START
n=node:index({key}='value')). Use the literal key names instead.
Using parameters to identify nodes in patterns
Parameters can not be used to identify nodes in a pattern (ie. MATCH ({node})-->(other))). Note that
this form was only possible when mixing the embedded Java API and Cypher, and thus does not
affect users of Neo4j Server.
Iteration syntax in FOREACH, EXTRACT, etc
The iterating functions use a | instead of a : to separate the components of the statement. For
example, EXTRACT(n in ns : n.prop) is replaced with EXTRACT(n in ns | n.prop). The iterating
functions include FOREACH, EXTRACT, REDUCE, ANY, ALL, SINGLE and NONE.
Alternative WITH syntax
The alternative WITH syntax, === <identifiers> ===, has been removed. Use the WITH keyword
instead.
The Reference Node
With the introduction of Labels in Neo4j 2.0 the Reference Node has become obsolete and
has been removed. Instead, labeled nodes has become the well-known starting points in your
graph. You can use a pattern like this to access a reference node: MATCH (ref:MyReference) RETURN
ref. Simply use one label per such starting point you want to use. Note that this should be
executed once during application initialization, to ensure that only a single reference node is
created per label. When migrating a database with an existing reference node, add a label to it
during migration, and then use it as per the previous pattern. This is how to add the label: START
ref=node(0) SET ref:MyReference. In case you have altered the database so a different node is the
reference node, substitute the node id in the statement.
Other significant changes
Plugins
Plugins are no longer distributed with Neo4j. Please see individual maintainers about availability.
For instance, the source for the Gremlin plugin will be available at: https://github.com/neo4jcontrib/gremlin-plugin
21.4.4.Upgrade 1.8 1.9

There are no store format changes between these versions so upgrading standalone instances simply
consists of starting the database with the newer version. In the case of High Availability (HA)
installations, the communication protocol and the master election algorithm have changed and a new
"rolling upgrade" feature has been added, removing the need to shut down the entire cluster. For more
information, refer to the "Upgrading a Neo4j HA Cluster" chapter of the HA section of the Neo4j
manual.
21.4.5.Upgrade 1.7 1.8

There are no store format changes between these versions so upgrading standalone instances simply
consists of starting the database with the newer version. In the case of High Availability (HA)
installations, the communication protocol and the master election algorithm have changed and a new
"rolling upgrade" feature has been added, removing the need to shut down the entire cluster. For more
information, refer to the "Upgrading a Neo4j HA Cluster" chapter of the HA section of the Neo4j
manual.
367
21.4.6.Upgrade 1.6 1.7

There are no store format changes between these versions, which means there is no particular
procedure you need to upgrade a single instance.
In an HA environment these steps need to be performed:
1.
2.
3.
4.
5.
shut down all the databases in the cluster

shut down the ZooKeeper cluster and clear the version-2 directories on all the ZooKeeper instances
start the ZooKeeper cluster again
remove the databases except the master and start the master database with 1.7
start up the other databases so that they get a copy from the master
21.4.7.Upgrade 1.5 1.6

This upgrade changes lucene version from 3.1 to 3.5. The upgrade itself is done by Lucene by loading
an index.
In an HA environment these steps need to be performed:
1.
2.
3.
4.
shut down all the databases in the cluster

shut down the ZooKeeper cluster and clear the version-2 directories on all the ZooKeeper instances
start the ZooKeeper cluster again
start up the other databases so that they get a copy from the master
21.4.8.Upgrade 1.4 1.5

This upgrade includes a significant change to the layout of property store files, which reduces their
size on disk, and improves IO performance. To achieve this layout change, the upgrade process takes
some time to process the whole of the existing database. You should budget for several minutes per
gigabyte of data as part of your upgrade planning.
Warning
The upgrade process for this upgrade temporarily requires additional disk space, for the
period while the upgrade is in progress. Before starting the upgrade to Neo4j 1.5, you
should ensure that the machine performing the upgrade has free space equal to the current
size of of the database on disk. You can find the current space occupied by the database by
inspecting the store file directory (data/graph.db is the default location in Neo4j server).
Once the upgrade is complete, this additional space is no longer required.
368
21.5.Setup for remote debugging

In order to configure the Neo4j server for remote debugging sessions, the Java debugging parameters
need to be passed to the Java process through the configuration. They live in the conf/neo4jwrapper.properties file.
In order to specify the parameters, add a line for the additional Java arguments like this:
Java Additional Parameters
wrapper.java.additional.1=-Dorg.neo4j.server.properties=conf/neo4j-server.properties
wrapper.java.additional.2=-Dlog4j.configuration=file:conf/log4j.properties
wrapper.java.additional.3=-agentlib:jdwp=transport=dt_socket,server=y,suspend=n,address=5005 \
-Xdebug-Xnoagent-Djava.compiler=NONE\
-Xrunjdwp:transport=dt_socket,server=y,suspend=n,address=5005
This configuration will start a Neo4j server ready for remote debugging attachement at localhost and
port 5005. Use these parameters to attach to the process from Eclipse, IntelliJ or your remote debugger
of choice after starting the server.
369
21.6.Usage Data Collector

The Neo4j Usage Data Collector is a sub-system that gathers usage data, reporting it to the UDCserver at udc.neo4j.org. It is easy to disable, and does not collect any data that is confidential. For
more information about what is being sent, see below.
The Neo4j team uses this information as a form of automatic, effortless feedback from the Neo4j
community. We want to verify that we are doing the right thing by matching download statistics with
usage statistics. After each release, we can see if there is a larger retention span of the server software.
The data collected is clearly stated here. If any future versions of this system collect additional data,
we will clearly announce those changes.
The Neo4j team is very concerned about your privacy. We do not disclose any personally identifiable
information.
21.6.1.Technical Information
To gather good statistics about Neo4j usage, UDC collects this information:
Kernel version: The build number, and if there are any modifications to the kernel.
Store id: A randomized globally unique id created at the same time a database is created.
Ping count: UDC holds an internal counter which is incremented for every ping, and reset for every
restart of the kernel.
Source: This is either "neo4j" or "maven". If you downloaded Neo4j from the Neo4j website, its
"neo4j", if you are using Maven to get Neo4j, it will be "maven".
Java version: The referrer string shows which version of Java is being used.
MAC address to uniquely identify instances behind firewalls.
Registration id: For registered server instances.
Tags about the execution context (e.g. test, language, web-container, app-container, spring, ejb).
Neo4j Edition (community, enterprise).
A hash of the current cluster name (if any).
Distribution information for Linux (rpm, dpkg, unknown).
User-Agent header for tracking usage of REST client drivers
After startup, UDC waits for ten minutes before sending the first ping. It does this for two reasons;
first, we dont want the startup to be slower because of UDC, and secondly, we want to keep pings
from automatic tests to a minimum. The ping to the UDC servers is done with a HTTP GET.
21.6.2.How to disable UDC

Weve tried to make it extremely easy to disable UDC. In fact, the code for UDC is not even included
in the kernel jar but as a completely separate component.
There are three ways you can disable UDC:
1. The easiest way is to just remove the neo4j-udc-*.jar file. By doing this, the kernel will not load
UDC, and no pings will be sent.
2. If you are using Maven, and want to make sure that UDC is never installed in your system, a
dependency element like this will do that:
<dependency>
<groupId>org.neo4j</groupId>
370

<artifactId>neo4j</artifactId>
<version>${neo4j-version}</version>
<exclusions>
<exclusion>
<artifactId>neo4j-udc</artifactId>
</exclusion>
</exclusions>
</dependency>
Where ${neo4j-version} is the Neo4j version in use.

3. Lastly, if you are using a packaged version of Neo4j, and do not want to make any change to
the jars, a system property setting like this will also make sure that UDC is never activated: Dneo4j.ext.udc.disable=true.
371
Chapter22.Configuration & Performance

In order to get optimum performance out of Neo4j for your application there are a few parameters that
can be tweaked. The two main components that can be configured are the Neo4j caches and the JVM
that Neo4j runs in. The following sections describe how to tune these.
372
Configuration & Performance
22.1.Introduction
To gain good performance, these are the things to look into first:
Make sure the JVM is not spending too much time performing garbage collection. Monitoring
heap usage on an application that uses Neo4j can be a bit confusing since Neo4j will increase the
size of caches if there is available memory and decrease if the heap is getting full. The goal is to
have a large enough heap to make sure that heavy/peak load will not result in so called GC trashing
(performance can drop as much as two orders of magnitude when GC trashing happens).
Start the JVM with the -server flag and a good sized heap (see Section22.8, JVM Settings).
Having too large heap may also hurt performance so you may have to try some different heap sizes.
Use the parallel/concurrent garbage collector (we found that -XX:+UseConcMarkSweepGC works well in
most use-cases)
22.1.1.How to add configuration settings

When creating the embedded Neo4j instance it is possible to pass in parameters contained in a
map where keys and values are strings, see the section called Starting an embedded database with
configuration settings for an example.
If no configuration is provided, the Database Kernel will try to determine suitable settings from the
information available via the JVM settings and the underlying operating system.
The JVM is configured by passing command line flags when starting the JVM. The most important
configuration parameters for Neo4j are the ones that control the memory and garbage collector, but
some of the parameters for configuring the Just In Time compiler are also of interest.
This is an example of starting up your applications main class using 64-bit server VM mode and a
heap space of 1GB:
java -d64 -server -Xmx1024m -cp /path/to/neo4j-kernel.jar:/path/to/jta.jar:/path/to/your-application.jar com.example.yourapp.MainClas
Looking at the example above you will also notice one of the most basic command line parameters:
the one for specifying the classpath. The classpath is the path in which the JVM searches for your
classes. It is usually a list of jar-files. Specifying the classpath is done by specifying the flag -cp (or classpath) and then the value of the classpath. For Neo4j applications this should at least include the
path to the Neo4j neo4j-kernel.jar and the Java Transaction API (jta.jar) as well as the path where
the classes for your application are located.
Tip
On Linux, Unix and Mac OS X each element in the path list are separated by a colon
symbol (:), on Windows the path elements are separated by a semicolon (;).
When using the Neo4j REST server, see Section22.2, Server Configuration for how to add
configuration settings for the database to the server.
373
22.2.Server Configuration
Quick info
The servers primary configuration file is found under conf/neo4j-server.properties

The conf/log4j.properties file contains the default server logging configuration
Low-level performance tuning parameters are found in conf/neo4j.properties
Configuraion of the deamonizing wrapper are found in conf/neo4j-wrapper.properties
HTTP logging configuration is found in conf/neo4j-http-logging.xml
22.2.1.Important server configurations parameters

The main configuration file for the server can be found at conf/neo4j-server.properties. This file
contains several important settings, and although the defaults are sensible administrators might choose
to make changes (especially to the port settings).
Set the location on disk of the database directory like this:
org.neo4j.server.database.location=data/graph.db
Note
On Windows systems, absolute locations including drive letters need to read "c:/data/db".
Specify the HTTP server port supporting data, administrative, and UI access:
org.neo4j.server.webserver.port=7474
Specify the client accept pattern for the webserver (default is 127.0.0.1, localhost only):
allow any client to connect
org.neo4j.server.webserver.address=0.0.0.0
For securing the Neo4j Server, see also Chapter25, Security

Set the location of the round-robin database directory which gathers metrics on the running server
instance:
org.neo4j.server.webadmin.rrdb.location=data/graph.db/../rrd
Set the URI path for the REST data API through which the database is accessed. This should be a
relative path.
org.neo4j.server.webadmin.data.uri=/db/data/
Setting the management URI for the administration API that the Webadmin tool uses. This should be
a relative path.
org.neo4j.server.webadmin.management.uri=/db/manage
Force the server to use IPv4 network addresses, in conf/neo4j-wrapper.conf under the section Java
Additional Parameters add a new paramter:
wrapper.java.additional.3=-Djava.net.preferIPv4Stack=true
Specify the number of threads used by the Neo4j Web server to control the level of concurrent HTTP
requests that the server will service.
374

org.neo4j.server.webserver.maxthreads=200
Note
The default value is 10 times the number of CPUs reported available by the JVM.
The server guards against orphaned transactions by using a timeout. If there are no requests for a
given transaction within the timeout period, the server will roll it back. You can configure the timeout
period by setting the following property to the number of seconds before timeout. The default timeout
is 60 seconds.
org.neo4j.server.transaction.timeout=60
Low-level performance tuning parameters can be explicitly set by referring to the following property:
org.neo4j.server.db.tuning.properties=neo4j.properties
If this property isnt set, the server will look for a file called neo4j.properties in the same directory as
the neo4j-server.properties file.
If this property isnt set, and there is no neo4j.properties file in the default configuration directory,
then the server will log a warning. Subsequently at runtime the database engine will attempt tune itself
based on the prevailing conditions.
22.2.2.Neo4j Database performance configuration

The fine-tuning of the low-level Neo4j graph database engine is specified in a separate properties file,
conf/neo4j.properties.
The graph database engine has a range of performance tuning options which are enumerated in
Section22.3, Server Performance Tuning. Note that other factors than Neo4j tuning should be
considered when performance tuning a server, including general server load, memory and file
contention, and even garbage collection penalties on the JVM, though such considerations are beyond
the scope of this configuration document.
22.2.3.Server logging configuration

Application events within Neo4j server are processed with java.util.logging <http://
download.oracle.com/javase/6/docs/technotes/guides/logging/overview.html> and configured in the file
conf/logging.properties.
By default it is setup to print INFO level messages both on screen and in a rolling file in data/log.
Most deployments will choose to use their own configuration here to meet local standards. During
development, much useful information can be found in the logs so some form of logging to disk is
well worth keeping. On the other hand, if you want to completely silence the console output, set:
java.util.logging.ConsoleHandler.level=OFF
By default log files are rotated at approximately 10Mb and named consecutively neo4j.<id>.<rotation
sequence #>.log To change the naming scheme, rotation frequency and backlog size modify
java.util.logging.FileHandler.pattern
java.util.logging.FileHandler.limit
java.util.logging.FileHandler.count
respectively to your needs. Details are available at the Javadoc for java.util.logging.FileHandler
<http://download.oracle.com/javase/6/docs/api/java/util/logging/FileHandler.html>.
Apart from log statements originating from the Neo4j server, other libraries report their messages
through various frameworks.
375
22.2.4.HTTP logging configuration

As well as logging events happening within the Neo4j server, it is possible to log the HTTP requests
and responses that the server consumes and produces. Configuring HTTP logging requires operators
to enable and configure the logger and where it will log; and then to optionally configure the log
format.
Warning
By default the HTTP logger uses Common Log Format <http://en.wikipedia.org/wiki/

Common_Log_Format> meaning that most Web server tooling can automtically consume such
logs. In general users should only enable HTTP logging, select an output directory, and if
necessary alter the rollover and retention policies.
To enable HTTP logging, edit the conf/neo4j-server.properties file to resemble the following:
org.neo4j.server.http.log.enabled=true
org.neo4j.server.http.log.config=conf/neo4j-http-logging.xml
org.neo4j.server.http.log.enabled=true tells the server that HTTP logging is enabled. HTTP logging
can be totally disabled by setting this property to false. org.neo4j.server.http.log.config=conf/neo4jhttp-logging.xml specifies the logging format and rollover policy file that governs how HTTP log
output is presented and archived. The defaults provided with Neo4j server uses an hourly log rotation
and Common Log Format <http://en.wikipedia.org/wiki/Common_Log_Format>.
If logging is set up to use log files then the server will check that the log file directory exists and is
writable. If this check fails, then the server will not startup and wil report the failure another available
channel like standard out.
22.2.5.Using X-Forwarded-Proto and X-Forwarded-Host to parameterize the

base URI for REST responses
There are occasions, for example when you want to host Neo4j server behind a proxy (e.g. one that
handles HTTPS traffic), and still have Neo4j respect the base URI of that externally visible proxy.
Ordinarily Neo4j uses the HOST header of the HTTP request to construct URIs in its responses. Where
a proxy is involved however, this is often undesirable. Instead Neo4j, uses the X-Forwarded-Host and
X-Forwarded-Proto headers provided by proxies to parameterize the URIs in the responses from the
databasess REST API. From the outside it looks as if the proxy generated that payload. If an XForwarded-Host header value contains more than one address (X-Forwarded-Host allows comma-andspace separated lists of addresses), Neo4j picks the first, which represents the client request.
In order to take advantage of this functionality, your proxy server must be configured to transmit these
headers to the Neo4j server. Failure to transmit both X-Forwarded-Host and X-Forwarded-Proto headers
will result in the original base URI being used.
22.2.6.Other configuration options

Enabling logging from the garbage collector
To get garbage collection logging output you have to pass the corresponding option to the server JVM
executable by setting in conf/neo4j-wrapper.conf the value
wrapper.java.additional.3=-Xloggc:data/log/neo4j-gc.log
This line is already present and needs uncommenting. Note also that logging is not directed to
console ; You will find the logging statements in data/log/ne4j-gc.log or whatever directory you set at
the option.
376

Disabling console types in Webadmin
You may, for security reasons, want to disable the the Neo4j Shell in Webadmin. Shells allow
arbitrary code execution, and so they could constitute a security risk if you do not trust all users of
your Neo4j Server.
In the conf/neo4j-server.properties file:
To disable all shells:
org.neo4j.server.manage.console_engines=
To enable only the Neo4j Shell:
org.neo4j.server.manage.console_engines=shell
377
22.3.Server Performance Tuning

At the heart of the Neo4j server is a regular Neo4j storage engine instance. That engine can be tuned
in the same way as the other embedded configurations, using the same file format. The only difference
is that the server must be told where to find the fine-tuning configuration.
Quick info
The neo4j.properties file is a standard configuration file that databases load in order to tune
their memory use and caching strategies.
See Section22.6, Caches in Neo4j for more information.
22.3.1.Specifying Neo4j tuning properties

The conf/neo4j-server.properties file in the server distribution, is the main configuration file for
the server. In this file we can specify a second properties file that contains the database tuning
settings (that is, the neo4j.properties file). This is done by setting a single property to point to a valid
neo4j.properties file:
org.neo4j.server.db.tuning.properties={neo4j.properties file}
On restarting the server the tuning enhancements specified in the neo4j.properties file will be loaded
and configured into the underlying database engine.
22.3.2.Specifying JVM tuning properties

Tuning the standalone server is achieved by editing the neo4j-wrapper.conf file in the conf directory of
NEO4J_HOME.
Edit the following properties:
neo4j-wrapper.conf JVM tuning properties
Property Name
Meaning
wrapper. java. initmemory
initial heap size (in MB)
wrapper. java. maxmemory
maximum heap size (in MB)
wrapper. java. additional. N
additional literal JVM parameter, where N is a

number for each
For more information on the tuning properties, see Section22.8, JVM Settings.
378
22.4.Performance Guide
This is the Neo4j performance guide. It will attempt to give you guidance on how to use Neo4j to
achieve maximum performance.
22.4.1.Try this first
The first thing is to make sure the JVM is running well and not spending too much time in garbage
collection. Monitoring heap usage of an application that uses Neo4j can be a bit confusing since Neo4j
will increase the size of caches if there is available memory and decrease if the heap is getting full.
The goal is to have a large enough heap so heavy/peak load will not result in so called GC trashing
(performance can drop as much as two orders of a magnitude when this happens).
Start the JVM with -server flag and -Xmx<good sized heap> (f.ex. -Xmx512M for 512Mb memory or
-Xmx3G for 3Gb memory). Having too large heap may also hurt performance so you may have to
try out some different heap sizes. Make sure a parallel/concurrent garbage collector is running (-XX:
+UseConcMarkSweepGC works well in most use-cases).
Finally make sure that the OS has some memory left to manage proper file system caches. This
means, if your server has 8GB of RAM dont use all of that RAM for heap (unless you have turned
off memory mapped buffers), but leave a good part of it to the OS. For more information on this see
Chapter22, Configuration & Performance.
For Linux specific tweaks, see Section22.12, Linux Performance Guide.
22.4.2.Neo4j primitives' lifecycle
Neo4j manages its primitives (nodes, relationships and properties) different depending on how you
use Neo4j. For example if you never get a property from a certain node or relationship that node or
relationship will not have its properties loaded into memory. The first time, after loading a node or
relationship, that any property is accessed all the properties are loaded for that entity. If any of those
properties contain an array larger than a few elements or a long string such values are loaded on
demand when requesting them individually. Similarly, relationships of a node will only be loaded the
first time they are requested for that node.
Nodes and relationships are cached using LRU caches. If you (for some strange reason) only work
with nodes the relationship cache will become smaller and smaller while the node cache is allowed
to grow (if needed). Working with many relationships and few nodes results in a bigger relationship
cache and smaller node cache.
The Neo4j API specification does not say anything about order regarding relationships so invoking
Node.getRelationships() may return the relationships in a different order than the previous invocation.
This allows us to make even heavier optimizations returning the relationships that are most commonly
traversed.
All in all Neo4j has been designed to be very adaptive depending on how it is used. The
(unachievable) overall goal is to be able to handle any incoming operation without having to go down
and work with the file/disk I/O layer.
22.4.3.Configuring Neo4j
In Chapter22, Configuration & Performance page theres information on how to configure Neo4j and
the JVM. These settings have a lot of impact on performance.
Disks, RAM and other tips

As always, as with any persistence solution, performance depends a lot on the persistence media used.
Better disks equals better performance.
379

If you have multiple disks or persistence media available it may be a good idea to split the store files
and transaction logs across those disks. Having the store files running on disks with low seek time can
do wonders for non-cached read operations. Today a typical mechanical drive has an average seek
time of about 5ms, this can cause a query or traversal to be very slow when the available amount of
RAM is too small or the configuration for caches and memory mapping is bad. A new good SATA
enabled SSD has an average seek time of <100 microseconds meaning those scenarios will execute at
least 50 times faster.
To avoid hitting disk you need more RAM. On a standard mechanical drive you can handle graphs
with a few tens of millions of primitives with 1-2GB of RAM. 4-8GB of RAM can handle graphs with
hundreds of millions of primitives while you need a good server with 16-32GB to handle billions of
primitives. However, if you invest in a good SSD you will be able to handle much larger graphs on
less RAM.
Use tools like vmstat or equivalent to gather information when your application is running. If you have
high I/O waits and not that many blocks going out/in to disks when running write/read transactions its
a sign that you need to tweak your Java heap, Neo4j cache and memory mapping settings (maybe even
get more RAM or better disks).
Write performance
If you are experiencing poor write performance after writing some data (initially fast, then massive
slowdown) it may be the operating system that is writing out dirty pages from the memory mapped
regions of the store files. These regions do not need to be written out to maintain consistency so to
achieve highest possible write speed that type of behavior should be avoided.
Another source of writes slowing down can be the transaction size. Many small transactions result
in a lot of I/O writes to disc and should be avoided. Too big transactions can result in OutOfMemory
errors, since the uncommitted transaction data is held on the Java Heap in memory. For details about
transaction management in Neo4j, please see Chapter16, Transaction Management.
The Neo4j kernel makes use of several store files and a logical log file to store the graph on disk. The
store files contain the actual graph and the log contains modifying operations. All writes to the logical
log are append-only and when a transaction is committed changes to the logical log will be forced
(fdatasync) down to disk. The store files are however not flushed to disk and writes to them are not
append-only either. They will be written to in a more or less random pattern (depending on graph
layout) and writes will not be forced to disk until the log is rotated or the Neo4j kernel is shut down.
Since random writes to memory mapped regions for the store files may happen it is very important
that the data does not get written out to disk unless needed. Some operating systems have very
aggressive settings regarding when to write out these dirty pages to disk. If the OS decides to
start writing out dirty pages of these memory mapped regions, write access to disk will stop being
sequential and become random. That hurts performance a lot, so to get maximum write performance
when using Neo4j make sure the OS is configured not to write out any of the dirty pages caused by
writes to the memory mapped regions of the store files. As an example, if the machine has 8GB of
RAM and the total size of the store files is 4GB (fully memory mapped) the OS has to be configured
to accept at least 50% dirty pages in virtual memory to make sure we do not get random disk writes.
Note
Make sure to read Section22.12, Linux Performance Guide as well for more specific
information.
380

Second level caching
While normally building applications and always assume the graph is in memory, sometimes it is
necessary to optimize certain performance critical sections. Neo4j adds a small overhead even if the
node, relationship or property in question is cached compared to in-memory data structures. If this
becomes an issue, use a profiler to find these hot spots and then add your own second-level caching.
We believe second-level caching should be avoided to greatest extent possible since it will force you
to take care of invalidation which sometimes can be hard. But when everything else fails you have to
use it so here is an example of how it can be done.
We have some POJO that wrapps a node holding its state. In this particular POJO we have overridden
the equals implementation.
public boolean equals( Object obj )
{
return underlyingNode.getProperty( "some_property" ).equals( obj );
}
public int hashCode()
{
return underlyingNode.getProperty( "some_property" ).hashCode();
}
This works fine in most scenarios, but in this particular scenario many instances of that POJO is
being worked with in nested loops adding/removing/getting/finding to collection classes. Profiling the
applications will show that the equals implementation is being called many times and can be viewed
as a hot spot. Adding second-level caching for the equals override will in this particular scenario
increase performance.
private Object cachedProperty = null;
public boolean equals( Object obj )
{
if ( cachedProperty == null )
{
cachedProperty = underlyingNode.getProperty( "some_property" );
}
return cachedProperty.equals( obj );
}
{
if ( cachedPropety == null )
{
cachedProperty = underlyingNode.getProperty( "some_property" );
}
return cachedProperty.hashCode();
}
The problem with this is that now we need to invalidate the cached property whenever some_property
is changed (may not be a problem in this scenario since the state picked for equals and hash code
computation often wont change).
Tip
To sum up, avoid second-level caching if possible and only add it when you really need it.
381
22.5.Kernel configuration
These are the configuration options you can pass to the neo4j kernel. They can either be passed as a
map when using the embedded database, or in the neo4j.properties file when using the Neo4j Server.
All stores total mapped memory size
all_stores_total_mapped_memory_size
The size to allocate for a memory mapping pool to be shared between all stores.
Default value: 524288000
Allow store upgrade
allow_store_upgrade
Whether to allow a store upgrade in case the current version of the database starts against an older
store version. Setting this to true does not guarantee successful upgrade, just that it allows an attempt
at it.
Default value: false
Array block size
array_block_size
Specifies the block size for storing arrays. This parameter is only honored when the store is created,
otherwise it is ignored. The default block size is 120 bytes, and the overhead of each block is the
same as for string blocks, i.e., 8 bytes.
Default value: 120
Backup slave
backup_slave
Mark this database as a backup slave.

Cache type
cache_type
The type of cache to use for nodes and relationships. Note that the Neo4j Enterprise Edition has the
additional 'hpc' cache type (High-Performance Cache). See the chapter on caches in the manual for
more information.
Default value: soft
Cypher parser version
cypher_parser_version
Enable this to specify a parser other than the default one.

Dump configuration
dump_configuration

382

Print out the effective Neo4j configuration after startup.
Forced kernel id
forced_kernel_id
An identifier that uniquely identifies this graph database instance within this JVM. Defaults to an
auto-generated number depending on how many instance are started in this JVM.
Gc monitor threshold
gc_monitor_threshold
The amount of time in ms the monitor thread has to be blocked before logging a message it was
blocked.
Default value: 200
Gc monitor wait time
gc_monitor_wait_time
Amount of time in ms the GC monitor thread will wait before taking another measurement.
Default value: 100
Intercept committing transactions
intercept_committing_transactions
Determines whether any TransactionInterceptors loaded will intercept prepared transactions before
they reach the logical log.
Intercept deserialized transactions
intercept_deserialized_transactions
Determines whether any TransactionInterceptors loaded will intercept externally received

transactions (e.g. in HA) before they reach the logical log and are applied to the store.
Keep logical logs
keep_logical_logs
Make Neo4j keep the logical transaction logs for being able to backup the database.Can be used for
specifying the threshold to prune logical logs after. For example "10 days" will prune logical logs
that only contains transactions older than 10 days from the current time, or "100k txs" will keep the
100k latest transactions and prune any older transactions.
Default value: true
Label block size
label_block_size
Default value: 60
383

Specifies the block size for storing labels exceeding in-lined space in node record. This parameter
is only honored when the store is created, otherwise it is ignored. The default block size is 60 bytes,
and the overhead of each block is the same as for string blocks, i.e., 8 bytes.
Default value: 60
Log mapped memory stats
log_mapped_memory_stats
Tell Neo4j to regularly log memory mapping statistics.

Log mapped memory stats filename
log_mapped_memory_stats_filename
The file where Neo4j will record memory mapping statistics.

Default value: mapped_memory_stats.log
Log mapped memory stats interval
log_mapped_memory_stats_interval
The number of records to be loaded between regular logging of memory mapping statistics.
Logging.threshold for rotation
logging. threshold_for_rotation
Threshold in bytes for when database logs (text logs, for debugging, that is) are rotated.
Logical log
logical_log
The base name for the logical log files, either an absolute path or relative to the store_dir setting.
This should generally not be changed.
Default value: nioneo_logical.log
Logical log rotation threshold
logical_log_rotation_threshold
Specifies at which file size the logical log will auto-rotate. 0 means that no rotation will
automatically occur based on file size. Default is 25M
Lucene searcher cache size
lucene_searcher_cache_size
Integer value that sets the maximum number of open lucene index searchers.
384

Mapped memory page size
mapped_memory_page_size
Target size for pages of mapped memory.

Neo store
neo_store
The base name for the Neo4j Store files, either an absolute path or relative to the store_dir setting.
This should generally not be changed.
Default value: neostore
Neostore.nodestore.db.mapped memory
neostore. nodestore. db. mapped_memory
The size to allocate for memory mapping the node store.

Neostore.propertystore.db.arrays.mapped memory
neostore. propertystore. db. arrays. mapped_memory
The size to allocate for memory mapping the array property store.
Neostore.propertystore.db.index.keys.mapped memory
neostore. propertystore. db. index. keys. mapped_memory
The size to allocate for memory mapping the store for property key strings.
Neostore.propertystore.db.index.mapped memory
neostore. propertystore. db. index. mapped_memory
The size to allocate for memory mapping the store for property key indexes.
Neostore.propertystore.db.mapped memory
neostore. propertystore. db. mapped_memory
The size to allocate for memory mapping the property value store.
Neostore.propertystore.db.strings.mapped memory
neostore. propertystore. db. strings. mapped_memory
The size to allocate for memory mapping the string property store.
385

Neostore.relationshipstore.db.mapped memory
neostore. relationshipstore. db. mapped_memory
The size to allocate for memory mapping the relationship store.

Node auto indexing
node_auto_indexing
Controls the auto indexing feature for nodes. Setting to false shuts it down, while true enables it by
default for properties listed in the node_keys_indexable setting.
Node keys indexable
node_keys_indexable
A list of property names (comma separated) that will be indexed by default. This applies to Nodes
only.
Query cache size
query_cache_size
Used to set the number of Cypher query execution plans that are cached.
Default value: 100
Read only database
read_only
Only allow read operations from this Neo4j instance. This mode still requires write access to the
directory for lock purposes
Rebuild idgenerators fast
rebuild_idgenerators_fast
Use a quick approach for rebuilding the ID generators. This give quicker recovery time, but will limit
the ability to reuse the space of deleted entities.
Default value: true
Relationship auto indexing
relationship_auto_indexing
Controls the auto indexing feature for relationships. Setting to false shuts it down, while true enables
it by default for properties listed in the relationship_keys_indexable setting.
Relationship grab size
relationship_grab_size
Default value: 100

386

How many relationships to read at a time during iteration
Default value: 100
Relationship keys indexable
relationship_keys_indexable
A list of property names (comma separated) that will be indexed by default. This applies to
Relationships only.
Remote logging enabled
remote_logging_enabled
Whether to enable logging to a remote server or not.

Remote logging host
remote_logging_host
Host for remote logging using LogBack SocketAppender.

Default value: 127.0.0.1
Remote logging port
remote_logging_port
Port for remote logging using LogBack SocketAppender.

Default value: 4560
Store dir
store_dir
The directory where the database files are located.

String block size
string_block_size
Specifies the block size for storing strings. This parameter is only honored when the store is created,
otherwise it is ignored. Note that each character in a string occupies two bytes, meaning that a block
size of 120 (the default size) will hold a 60 character long string before overflowing into a second
block. Also note that each block carries an overhead of 8 bytes. This means that if the block size is
120, the size of the stored records will be 128 bytes.
Default value: 120
Tx manager impl
tx_manager_impl
The name of the Transaction Manager service to use as defined in the TM service provider
constructor.
Default value: native
387

Use memory mapped buffers
use_memory_mapped_buffers
Tell Neo4j to use memory mapped buffers for accessing the native storage layer.
Default value: true
388
22.6.Caches in Neo4j
For how to provide custom configuration to Neo4j, see Section22.1, Introduction.
Neo4j utilizes two different types of caches: A file buffer cache and an object cache. The file buffer
cache caches the storage file data in the same format as it is stored on the durable storage media.
The object cache caches the nodes, relationships and properties in a format that is optimized for high
traversal speeds and transactional writes.
22.6.1.File buffer cache

Quick info
The file buffer cache is sometimes called low level cache or file system cache.
It caches the Neo4j data as stored on the durable media.
It uses the operating system memory mapping features when possible.
Neo4j will configure the cache automatically as long as the heap size of the JVM is
configured properly.
The file buffer cache caches the Neo4j data in the same format as it is represented on the durable
storage media. The purpose of this cache layer is to improve both read and write performance. The
file buffer cache improves write performance by writing to the cache and deferring durable write until
the logical log is rotated. This behavior is safe since all transactions are always durably written to the
logical log, which can be used to recover the store files in the event of a crash.
Since the operation of the cache is tightly related to the data it stores, a short description of the Neo4j
durable representation format is necessary background. Neo4j stores data in multiple files and relies
on the underlying file system to handle this efficiently. Each Neo4j storage file contains uniform fixed
size records of a particular type:
Store file
Record size Contents
neostore.nodestore.db
14 B
Nodes
neostore.relationshipstore.db
33 B
Relationships
neostore.propertystore.db
41 B
Properties for nodes and

relationships
neostore.propertystore.db.strings
128 B
Values of string properties
neostore.propertystore.db.arrays
128 B
Values of array properties
For strings and arrays, where data can be of variable length, data is stored in one or more 120B
chunks, with 8B record overhead. The sizes of these blocks can actually be configured when the store
is created using the string_block_size and array_block_size parameters. The size of each record type
can also be used to calculate the storage requirements of a Neo4j graph or the appropriate cache size
for each file buffer cache. Note that some strings and arrays can be stored without using the string
store or the array store respectively, see Section22.9, Compressed storage of short strings and
Section22.10, Compressed storage of short arrays.
Neo4j uses multiple file buffer caches, one for each different storage file. Each file buffer cache
divides its storage file into a number of equally sized windows. Each cache window contains an even
number of storage records. The cache holds the most active cache windows in memory and tracks hit
389

vs. miss ratio for the windows. When the hit ratio of an uncached window gets higher than the miss
ratio of a cached window, the cached window gets evicted and the previously uncached window is
cached instead.
Important
Note that the block sizes can only be configured at store creation time.
Configuration
Parameter
Possible values
use_memory_mapped_buffers
true
or false
Effect
If set to true Neo4j will use
the operating systems memory
mapping functionality for the
file buffer cache windows. If set
to false Neo4j will use its own
buffer implementation. In this case
the buffers will reside in the JVM
heap which needs to be increased
accordingly. The default value for
this parameter is true, except on
Windows.
neostore. nodestore. db.

mapped_memory
The maximum amount of memory

to use for the file buffer cache of
the node storage file.
neostore. relationshipstore. db.

mapped_memory

the relationship store file.
neostore. propertystore. db. index.

keys. mapped_memory
neostore. propertystore. db. index.
mapped_memory
neostore. propertystore. db.
mapped_memory

The maximum amount of to use for the file buffer cache of
memory to use for memory the something-something file.
mapped buffers for this file The maximum amount of memory
buffer cache. The default to use for the file buffer cache of
unit is MiB, for other units the something-something file.
use any of the following
suffixes: B, k, M or G.
the property storage file.
neostore. propertystore. db.

strings. mapped_memory

the string property storage file.
neostore. propertystore. db. arrays.

mapped_memory

the array property storage file.
string_block_size
The number of
bytes per block.
390
Specifies the block size for storing

strings. This parameter is only
honored when the store is created,
otherwise it is ignored. Note that

Parameter
Possible values

arrays. This parameter is only
honored when the store is created,
otherwise it is ignored. The default
block size is 120 bytes, and the
overhead of each block is the same
as for string blocks, i.e., 8 bytes.
array_block_size
dump_configuration
Effect
each character in a string occupies
two bytes, meaning that a block
size of 120 (the default size) will
hold a 60 character long string
before overflowing into a second
block. Also note that each block
carries an overhead of 8 bytes.
This means that if the block size is
120, the size of the stored records
will be 128 bytes.
true
or false
If set to true the current

configuration settings will be
written to the default system
output, mostly the console or the
logfiles.
When memory mapped buffers are used (use_memory_mapped_buffers = true) the heap size of the JVM
must be smaller than the total available memory of the computer, minus the total amount of memory
used for the buffers. When heap buffers are used (use_memory_mapped_buffers = false) the heap size of
the JVM must be large enough to contain all the buffers, plus the runtime heap memory requirements
of the application and the object cache.
When reading the configuration parameters on startup Neo4j will automatically configure the
parameters that are not specified. The cache sizes will be configured based on the available memory
on the computer, how much is used by the JVM heap, and how large the storage files are.
22.6.2.Object cache
Quick info
The object cache is sometimes called high level cache.
It caches the Neo4j data in a form optimized for fast traversal.
The object cache caches individual nodes and relationships and their properties in a form that is
optimized for fast traversal of the graph. There are two different categories of object caches in Neo4j.
There is the reference caches. Here Neo4j will utilize as much as it can out of the allocated heap
memory for the JVM for object caching and relies on garbage collection for eviction from the cache
in an LRU manner. Note however that Neo4j is competing for the heap space with other objects
in the same JVM, such as a your application, if deployed in embedded mode, and Neo4j will let the
application win by using less memory if the application needs more.
391
Note
The High-Performance Cache described below is only available in the Neo4j Enterprise
Edition.
The other is the High-Performance Cache which gets assigned a certain amount of space in the JVM
heap and will purge objects whenever it grows bigger than that. It is assigned a maximum amount
of memory which the sum of all cached objects in it will not exceed. Objects will be evicted from
cache when the maximum size is about to be reached, instead of relying on garbage collection (GC)
to make that decision. Here the competition with other objects in the heap as well as GC-pauses can
be better controlled since the cache gets assigned a maximum heap space usage. The overhead of the
High-Performance Cache is also much smaller as well as insert/lookup times faster than for reference
caches.
Tip
The use of heap memory is subject to the java garbage collector depending on the cache
type some tuning might be needed to play well with the GC at large heap sizes. Therefore,
assigning a large heap for Neo4js sake isnt always the best strategy as it may lead to long
GC-pauses. Instead leave some space for Neo4js filesystem caches. These are outside of
the heap and under under the kernels direct control, thus more efficiently managed.
The content of this cache are objects with a representation geared towards supporting the Neo4j object
API and graph traversals. Reading from this cache is 5 to 10 times faster than reading from the file
buffer cache. This cache is contained in the heap of the JVM and the size is adapted to the current
amount of available heap memory.
Nodes and relationships are added to the object cache as soon as they are accessed. The cached
objects are however populated lazily. The properties for a node or relationship are not loaded until
properties are accessed for that node or relationship. String (and array) properties are not loaded until
that particular property is accessed. The relationships for a particular node is also not loaded until the
relationships are accessed for that node.
Configuration
The main configuration parameter for the object cache is the cache_type parameter. This specifies
which cache implementation to use for the object cache. Note that there will exist two cache instances,
one for nodes and one for relationships. The available cache types are:
cache_type
Description
none
Do not use a high level cache. No objects will be cached.
soft
Provides optimal utilization of the available memory. Suitable for high performance
traversal. May run into GC issues under high load if the frequently accessed parts of
the graph does not fit in the cache.
This is the default cache implementation.
weak
Provides short life span for cached objects. Suitable for high throughput applications
where a larger portion of the graph than what can fit into memory is frequently
accessed.
strong
This cache will hold on to all data that gets loaded to never release it again. Provides
good performance if your graph is small enough to fit in memory.
hpc
The High-Performance Cache. Provides means of assigning a specific amount of

memory to dedicate to caching loaded nodes and relationships. Small footprint and
392

cache_type
Description
fast insert/lookup. Should be the best option for most scenarios. See below on how to
configure it. Note that this option is only available in the Neo4j Enterprise Edition.
High-Performance Cache
Since the High-Performance Cache operates with a maximum size in the JVM it may be configured
per use case for optimal performance. There are two aspects of the cache size.
One is the size of the array referencing the objects that are put in the cache. It is specified as a
fraction of the heap, for example specifying 5 will let that array itself take up 5% out of the entire
heap. Increasing this figure (up to a maximum of 10) will reduce the chance of hash collisions at the
expense of more heap used for it. More collisions means more redundant loading of objects from the
low level cache.
configuration option
Description (what it controls)
Example
value
node_cache_array_fraction
Fraction of the heap to dedicate to the array holding

the nodes in the cache (max 10).
relationship_cache_array_fraction
Fraction of the heap to dedicate to the array holding

the relationships in the cache (max 10).
The other aspect is the maximum size of all the objects in the cache. It is specified as size in bytes, for
example 500M for 500 megabytes or 2G for two gigabytes. Right before the maximum size is reached
a purge is performed where (currently) random objects are evicted from the cache until the cache size
gets below 90% of the maximum size. Optimal settings for the maximum size depends on the size
of your graph. The configured maximum size should leave enough room for other objects to coexist
in the same JVM, but at the same time large enough to keep loading from the low level cache at a
minimum. Predicted load on the JVM as well as layout of domain level objects should also be take
into consideration.
configuration option
Description (what it controls)
Example
value
node_cache_size
Maximum size of the heap memory to dedicate to the cached

nodes.
2G
relationship_cache_size
Maximum size of the heap memory to dedicate to the cached

relationships.
800M
You can read about references and relevant JVM settings for Sun HotSpot here:
Understanding soft/weak references <http://weblogs.java.net/blog/enicholas/archive/2006/05/
understanding_w.html>
How Hotspot Decides to Clear SoftReferences <http://jeremymanson.blogspot.com/2009/07/howhotspot-decides-to-clear_07.html>
HotSpot FAQ <http://www.oracle.com/technetwork/java/hotspotfaq-138619.html#gc_softrefs>
393
22.7.Logical logs
Logical logs in Neo4j are the journal of which operations happens and are the source of truth in
scenarios where the database needs to be recovered after a crash or similar. Logs are rotated every
now and then (defaults to when they surpass 25 Mb in size) and the amount of legacy logs to keep can
be configured. Purpose of keeping a history of logical logs include being able to serve incremental
backups as well as keeping an HA cluster running.
For any given configuration at least the latest non-empty logical log will be kept, but configuration
can be supplied to control how much more to keep. There are several different means of controlling it
and the format in which configuration is supplied is:
keep_logical_logs=<true/false>
keep_logical_logs=<amount> <type>
For example:
Will keep logical logs indefinitely
keep_logical_logs=true
Will keep only the most recent non-empty log
keep_logical_logs=false
Will keep logical logs which contains any transaction committed within 30 days
keep_logical_logs=30 days
Will keep logical logs which contains any of the most recent 500 000 transactions
keep_logical_logs=500k txs
Full list:
Type
Description
Example
files
Number of most recent logical log files to

keep
"10 files"
size
Max disk size to allow log files to occupy
"300M size" or
"1G size"
txs
Number of latest transactions to keep Keep
"250k txs" or
"5M txs"
hours
Keep logs which contains any transaction

"10 hours"
committed within N hours from current time
days
Keep logs which contains any transaction

committed within N days from current time
394
"50 days"
22.8.JVM Settings
22.8.1.Background
There are two main memory parameters for the JVM, one controls the heap space and the other
controls the stack space. The heap space parameter is the most important one for Neo4j, since this
governs how many objects you can allocate. The stack space parameter governs the how deep the call
stack of your application is allowed to get.
When it comes to heap space the general rule is: the larger heap space you have the better, but make
sure the heap fits in the RAM memory of the computer. If the heap is paged out to disk performance
will degrade rapidly. Having a heap that is much larger than what your application needs is not
good either, since this means that the JVM will accumulate a lot of dead objects before the garbage
collector is executed, this leads to long garbage collection pauses and undesired performance
behavior.
Having a larger heap space will mean that Neo4j can handle larger transactions and more concurrent
transactions. A large heap space will also make Neo4j run faster since it means Neo4j can fit a larger
portion of the graph in its caches, meaning that the nodes and relationships your application uses
frequently are always available quickly. The default heap size for a 32bit JVM is 64MB (and 30%
larger for 64bit), which is too small for most real applications.
Neo4j works fine with the default stack space configuration, but if your application implements some
recursive behavior it is a good idea to increment the stack size. Note that the stack size is shared for
all threads, so if you application is running a lot of concurrent threads it is a good idea to increase the
stack size.
The heap size is set by specifying the -Xmx???m parameter to hotspot, where ??? is the heap size in
megabytes. Default heap size is 64MB for 32bit JVMs, 30% larger (appr. 83MB) for 64bit JVMs.
The stack size is set by specifying the -Xss???m parameter to hotspot, where ??? is the stack size in
megabytes. Default stack size is 512kB for 32bit JVMs on Solaris, 320kB for 32bit JVMs on Linux
(and Windows), and 1024kB for 64bit JVMs.
Most modern CPUs implement a Non-Uniform Memory Access (NUMA) architecture <http://
en.wikipedia.org/wiki/Non-Uniform_Memory_Access>, where different parts of the memory
have different access speeds. Suns Hotspot JVM is able to allocate objects with awareness of the
NUMA structure as of version 1.6.0 update 18. When enabled this can give up to 40% performance
improvements. To enabled the NUMA awareness, specify the -XX:+UseNUMA parameter (works only
when using the Parallel Scavenger garbage collector (default or -XX:+UseParallelGC not the concurrent
mark and sweep one).
Properly configuring memory utilization of the JVM is crucial for optimal performance. As an
example, a poorly configured JVM could spend all CPU time performing garbage collection (blocking
all threads from performing any work). Requirements such as latency, total throughput and available
hardware have to be considered to find the right setup. In production, Neo4j should run on a multi
core/CPU platform with the JVM in server mode.
22.8.2.Configuring heap size and GC

A large heap allows for larger node and relationship caches which is a good thing but large
heaps can also lead to latency problems caused by full garbage collection. The different high level
cache implementations available in Neo4j together with a suitable JVM configuration of heap size and
garbage collection (GC) should be able to handle most workloads.
395

The default cache (soft reference based LRU cache) works best with a heap that never gets full: a
graph where the most used nodes and relationships can be cached. If the heap gets too full there is a
risk that a full GC will be triggered; the larger the heap, the longer it can take to determine what soft
references should be cleared.
Using the strong reference cache means that all the nodes and relationships being used must fit in the
available heap. Otherwise there is a risk of getting out-of-memory exceptions. The soft reference and
strong reference caches are well suited for applications were the overal throughput is important.
The weak reference cache basically needs enough heap to handle the peak load of the
application peak load multiplied by the average memory required per request. It is well suited for
low latency requirements were GC interuptions are not acceptable.
Important
When running Neo4j on Windows, keep in mind that the memory mapped buffers are
allocated on heap by default, so they need to be taken into account when determining heap
size.
Guidelines for heap size
Number of
RAM size
primitives
Heap
configuration
Reserved RAM
for the OS
10M
2GB
512MB
the rest
100M
8GB+
1-4GB
1-2GB
1B+
16GB-32GB+
4GB+
1-2GB
Tip
The recommended garbage collector to use when running Neo4j in production is the
Concurrent Mark and Sweep Compactor turned on by supplying -XX:+UseConcMarkSweepGC
as a JVM parameter.
When having made sure that the heap size is well configured the second thing to tune in order to tune
the garbage collector for your application is to specify the sizes of the different generations of the
heap. The default settings are well tuned for "normal" applications, and work quite well for most
applications, but if you have an application with either really high allocation rate, or a lot of long
lived objects you might want to consider tuning the sizes of the heap generation. The ratio between
the young and tenured generation of the heap is specified by using the -XX:NewRatio= command line
option (where is replaced by a number). The default ratio is 1:12 for client mode JVM, and 1:8 for
server mode JVM. You can also specify the size of the young generation explicitly using the -Xmn
command line option, which works just like the -Xmx option that specifies the total heap space.
GC shortname
Generation
Command line
parameter
Comment
Copy
Young
-XX:+UseSerialGC
The Copying collector
MarkSweepCompact
Tenured
-XX:+UseSerialGC
The Mark and Sweep

Compactor
ConcurrentMarkSweep Tenured
-XX:+UseConcMarkSweepGC
The Concurrent Mark

and Sweep Compactor
ParNew
-XX:+UseParNewGC
The parallel
Young Generation
Young
396

GC shortname
Generation
Command line
parameter
Comment
Collector can only
be used with the
Concurrent mark and
sweep compactor.
PS Scavenge
Young
-XX:+UseParallelGC
The parallel object

scavenger
PS MarkSweep
Tenured
-XX:+UseParallelGC
The parallel mark and

sweep collector
These are the default configurations on some platforms according to our non-exhaustive research:
JVM
-d32 -client
-d32 -server
Mac OS X Snow
Leopard, 64-bit,
Hotspot 1.6.0_17
ParNew and
PS Scavenge
ConcurrentMarkSweep MarkSweep
Ubuntu, 32-bit,
Hotspot 1.6.0_16
Copy and
MarkSweepCompact
-d64 -server
and PS ParNew and
Copy and
MarkSweepCompact
397
-d64 -client
PS Scavenge
ConcurrentMarkSweep MarkSweep
N/A
N/A
and PS
22.9.Compressed storage of short strings
Neo4j will try to classify your strings in a short string class and if it manages that it will treat it
accordingly. In that case, it will be stored without indirection in the property store, inlining it instead
in the property record, meaning that the dynamic string store will not be involved in storing that value,
leading to reduced disk footprint. Additionally, when no string record is needed to store the property,
it can be read and written in a single lookup, leading to performance improvements and less disk space
required.
The various classes for short strings are:
Numerical, consisting of digits 0..9 and the punctuation space, period, dash, plus, comma and
apostrophe.
Date, consisting of digits 0..9 and the punctuation space dash, colon, slash, plus and comma.
Hex (lower case), consisting of digits 0..9 and lower case letters a..f
Hex (upper case), consisting of digits 0..9 and upper case letters a..f
Upper case, consisting of upper case letters A..Z, and the punctuation space, underscore, period,
dash, colon and slash.
Lower case, like upper but with lower case letters a..z instead of upper case
E-mail, consisting of lower case letters a..z and the punctuation comma, underscore, period, dash,
plus and the at sign (@).
URI, consisting of lower case letters a..z, digits 0..9 and most punctuation available.
Alpha-numerical, consisting of both upper and lower case letters a..zA..z, digits 0..9 and
punctuation space and underscore.
Alpha-symbolical, consisting of both upper and lower case letters a..zA..Z and the punctuation
space, underscore, period, dash, colon, slash, plus, comma, apostrophe, at sign, pipe and semicolon.
European, consisting of most accented european characters and digits plus punctuation space, dash,
underscore and period like latin1 but with less punctuation.
Latin 1.
UTF-8.
In addition to the strings contents, the number of characters also determines if the string can be
inlined or not. Each class has its own character count limits, which are
Character count limits
String class
Character
count
limit
Numerical, Date and Hex
54
Uppercase, Lowercase and E-mail
43
URI, Alphanumerical and

Alphasymbolical
36
European
31
Latin1
27
UTF-8
14
That means that the largest inline-able string is 54 characters long and must be of the Numerical class
and also that all Strings of size 14 or less will always be inlined.
398

Also note that the above limits are for the default 41 byte PropertyRecord layout if that parameter
is changed via editing the source and recompiling, the above have to be recalculated.
399
22.10.Compressed storage of short arrays

Neo4j will try to store your primitive arrays in a compressed way, so as to save disk space and
possibly an I/O operation. To do that, it employs a "bit-shaving" algorithm that tries to reduce the
number of bits required for storing the members of the array. In particular:
1.
2.
3.
4.
For each member of the array, it determines the position of leftmost set bit.
Determines the largest such position among all members of the array
It reduces all members to that number of bits
Stores those values, prefixed by a small header.
That means that when even a single negative value is included in the array then the natural size of the
primitives will be used.
There is a possibility that the result can be inlined in the property record if:
It is less than 24 bytes after compression
It has less than 64 members
For example, an array long[] {0L, 1L, 2L, 4L} will be inlined, as the largest entry (4) will require
3 bits to store so the whole array will be stored in 4*3=12 bits. The array long[] {-1L, 1L, 2L, 4L}
however will require the whole 64 bits for the -1 entry so it needs 64*4 = 32 bytes and it will end up
in the dynamic store.
400
22.11.Memory mapped IO settings

22.11.1.Introduction
Each file in the Neo4j store can use memory mapped I/O for reading/writing. Best performance is
achieved if the full file can be memory mapped but if there isnt enough memory for that Neo4j will
try and make the best use of the memory it gets (regions of the file that get accessed often will more
likely be memory mapped).
Important
Neo4j makes heavy use of the java.nio package. Native I/O will result in memory being
allocated outside the normal Java heap so that memory usage needs to be taken into
consideration. Other processes running on the OS will impact the availability of such
memory. Neo4j will require all of the heap memory of the JVM plus the memory to be
used for memory mapping to be available as physical memory. Other processes may thus
not use more than what is available after the configured memory allocation is made for
Neo4j.
A well configured OS with large disk caches will help a lot once we get cache misses in the node and
relationship caches. Therefore it is not a good idea to use all available memory as Java heap.
If you look into the directory of your Neo4j database, you will find its store files, all prefixed by
neostore:
nodestore stores information about nodes
relationshipstore holds all the relationships
propertystore stores information of properties and all simple properties such as primitive types
(both for relationships and nodes)
propertystore strings stores all string properties
propertystore arrays stores all array properties
There are other files there as well, but they are normally not interesting in this context.
This is how the default memory mapping configuration looks:
neostore.nodestore.db.mapped_memory=25M
neostore.relationshipstore.db.mapped_memory=50M
neostore.propertystore.db.mapped_memory=90M
neostore.propertystore.db.strings.mapped_memory=130M
neostore.propertystore.db.arrays.mapped_memory=130M
22.11.2.Optimizing for traversal speed example

To tune the memory mapping settings start by investigating the size of the different store files found
in the directory of your Neo4j database. Here is an example of some of the files and sizes in a Neo4j
database:
14M neostore.nodestore.db
510M neostore.propertystore.db
1.2G neostore.propertystore.db.strings
304M neostore.relationshipstore.db
In this example the application is running on a machine with 4GB of RAM. Weve reserved about
2GB for the OS and other programs. The Java heap is set to 1.5GB, that leaves about 500MB of RAM
that can be used for memory mapping.
401
Tip
If traversal speed is the highest priority it is good to memory map as much as possible of
the node- and relationship stores.
An example configuration on the example machine focusing on traversal speed would then look
something like:
neostore.relationshipstore.db.mapped_memory=285M
22.11.3.Batch insert example

Read general information on batch insertion in Chapter35, Batch Insertion.
The configuration should suit the data set you are about to inject using BatchInsert. Lets say we
have a random-like graph with 10M nodes and 100M relationships. Each node (and maybe some
relationships) have different properties of string and Java primitive types (but no arrays). The
important thing with a random graph will be to give lots of memory to the relationship and node store:
neostore.relationshipstore.db.mapped_memory=3G
The configuration above will fit the entire graph (with exception to properties) in memory.
A rough formula to calculate the memory needed for the nodes:
number_of_nodes * 9 bytes
and for relationships:

number_of_relationships * 33 bytes
Properties will typically only be injected once and never read so a few megabytes for the property
store and string store is usually enough. If you have very large strings or arrays you may want to
increase the amount of memory assigned to the string and array store files.
An important thing to remember is that the above configuration will need a Java heap of 3.3G+ since
in batch inserter mode normal Java buffers that gets allocated on the heap will be used instead of
memory mapped ones.
402
22.12.Linux Performance Guide

22.12.1.Introduction
The key to achieve good performance on reads and writes is to have lots of RAM since disks are so
slow. This guide will focus on achieving good write performance on a Linux kernel based operating
system.
If you have not already read the information available in Chapter22, Configuration & Performance,
do that now to get some basic knowledge on memory mapping and store files with Neo4j.
This section will guide you through how to set up a file system benchmark and use it to configure
your system in a better way.
22.12.2.File system benchmark

Setup
Create a large file with random data. The file should fit in RAM so if your machine has 4GB of RAM
a 1-2GB file with random data will be enough. After the file has been created we will read the file
sequentially a few times to make sure it is cached.
$ dd if=/dev/urandom of=store bs=1M count=1000
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 263.53 s, 4.0 MB/s
$
$ dd if=store of=/dev/null bs=100M
10+0 records in
10+0 records out
1048576000 bytes (1.0 GB) copied, 38.6809 s, 27.1 MB/s
$
10+0 records in
10+0 records out
1048576000 bytes (1.0 GB) copied, 1.52365 s, 688 MB/s
10+0 records in
10+0 records out
1048576000 bytes (1.0 GB) copied, 0.776044 s, 1.4 GB/s
If you have a standard hard drive in the machine you may know that it is not capable of transfer
speeds as high as 1.4GB/s. What is measured is how fast we can read a file that is cached for us by the
operating system.
Next we will use a small utility that simulates the Neo4j kernel behavior to benchmark write speed of
the system.
$ git clone git@github.com:neo4j-contrib/tooling.git

...
$ cd tooling/write-test/
$ mvn compile
[INFO] Scanning for projects...
...
$ ./run
Usage: <large file> <log file> <[record size] [min tx size] [max tx size] [tx count] <[--nosync | --nowritelog | --nowritestore | --n
The utility will be given a store file (large file we just created) and a name of a log file. Then a record
size in bytes, min tx size, max tx size and transaction count must be set. When started the utility will
map the large store file entirely in memory and read (transaction size) records from it randomly and
403

then write them sequentially to the log file. The log file will then force changes to disk and finally the
records will be written back to the store file.
Running the benchmark
Lets try to benchmark 100 transactions of size 100-500 with a record size of 33 bytes (same record
size used by the relationship store).
$ ./run store logfile 33 100 500 100
tx_count[100] records[30759] fdatasyncs[100] read[0.96802425 MB] wrote[1.9360485 MB]
Time was: 4.973
20.108585 tx/s, 6185.2 records/s, 20.108585 fdatasyncs/s, 199.32773 kB/s on reads, 398.65546 kB/s on writes
We see that we get about 6185 record updates/s and 20 transactions/s with the current transaction
size. We can change the transaction size to be bigger, for example writing 10 transactions of size
1000-5000 records:
Time was: 0.792
With larger transaction we will do fewer of them per second but record throughput will increase. Lets
see if it scales, 10 transactions in under 1s then 100 of them should execute in about 10s:
Time was: 65.115
This is not very linear scaling. We modified a bit more than 10x records in total but the time jumped
up almost 100x. Running the benchmark watching vmstat output will reveal that something is not as it
should be:
$ vmstat 3
procs -----------memory---------- ---swap-- -----io---- -system-- ----cpu---r b
swpd
free
buff cache
si
so
bi
bo
in
cs us sy id wa
0 1 47660 298884 136036 2650324
0
0
0 10239 1167 2268 5 7 46 42
0 1 47660 302728 136044 2646060
0
0
0 7389 1267 2627 6 7 47 40
0 1 47660 302408 136044 2646024
0
0
0 11707 1861 2016 8 5 48 39
0 2 47660 302472 136060 2646432
0
0
0 10011 1704 1878 4 7 49 40
0 1 47660 303420 136068 2645788
0
0
0 13807 1406 1601 4 5 44 47
There are a lot of blocks going out to IO, way more than expected for the write speed we are seeing in
the benchmark. Another observation that can be made is that the Linux kernel has spawned a process
called "flush-x:x" (run top) that seems to be consuming a lot of resources.
The problem here is that the Linux kernel is trying to be smart and write out dirty pages from the
virtual memory. As the benchmark will memory map a 1GB file and do random writes it is likely that
this will result in 1/4 of the memory pages available on the system to be marked as dirty. The Neo4j
kernel is not sending any system calls to the Linux kernel to write out these pages to disk however the
Linux kernel decided to start doing so and it is a very bad decision. The result is that instead of doing
sequential like writes down to disk (the logical log file) we are now doing random writes writing
regions of the memory mapped file to disk.
It is possible to observe this behavior in more detail by looking at /proc/vmstat "nr_dirty" and
"nr_writeback" values. By default the Linux kernel will start writing out pages at a very low ratio of
dirty pages (10%).
$ sync
$ watch grep -A 1 dirty /proc/vmstat
...
404

nr_dirty 22
nr_writeback 0
The "sync" command will write out all data (that needs writing) from memory to disk. The second
command will watch the "nr_dirty" and "nr_writeback" count from vmstat. Now start the benchmark
again and observe the numbers:
nr_dirty 124947
nr_writeback 232
The "nr_dirty" pages will quickly start to rise and after a while the "nr_writeback" will also increase
meaning the Linux kernel is scheduling a lot of pages to write out to disk.
Fixing the problem
As we have 4GB RAM on the machine and memory map a 1GB file that does not need its content
written to disk (until we tell it to do so because of logical log rotation or Neo4j kernel shutdown) it
should be possible to do endless random writes to that memory with high throughput. All we have to
do is to tell the Linux kernel to stop trying to be smart. Edit the /etc/sysctl.conf (need root access) and
add the following lines:
vm.dirty_background_ratio = 50
vm.dirty_ratio = 80
Then (as root) execute:

sysctl -p
The "vm.dirty_background_ratio" tells at what ratio should the linux kernel start the background task
of writing out dirty pages. We increased this from the default 10% to 50% and that should cover the
1GB memory mapped file. The "vm.dirty_ratio" tells at what ratio all IO writes become synchronous,
meaning that we can not do IO calls without waiting for the underlying device to complete them
(which is something you never want to happen).
Rerun the benchmark:
Time was: 6.781
Results are now more in line with what can be expected, 10x more records modified results in 10x
longer execution time. The vmstat utility will not report any absurd amount of IO blocks going out (it
reports the ones caused by the fdatasync to the logical log) and Linux kernel will not spawn a "flushx:x" background process writing out dirty pages caused by writes to the memory mapped store file.
22.12.3.File system tuning for high IO

In order to support the high IO load of small transactions from a database, the underlying file system
should be tuned. Symptoms for this are low CPU load with high iowait. In this case, there are a couple
of tweaks possible on Linux systems:
Disable access-time updates: noatime,nodiratime flags for disk mount command or in the /etc/fstab
for the database disk volume mount.
Tune the IO scheduler for high disk IO on the database disk.
22.12.4.Setting the number of open files

Linux platforms impose an upper limit on the number of concurrent files a user may have open. This
number is reported for the current user and session with the command
405

user@localhost:~$ ulimit -n
1024
The usual default of 1024 is often not enough, especially when many indexes are used or a server
installation sees too many connections (network sockets count against that limit as well). Users are
therefore encouraged to increase that limit to a healthy value of 40000 or more, depending on usage
patterns. Setting this value via the ulimit command is possible only for the root user and that for that
session only. To set the value system wide you have to follow the instructions for your platform.
What follows is the procedure to set the open file descriptor limit to 40k for user neo4j under Ubuntu
10.04 and later. If you opted to run the neo4j service as a different user, change the first field in step 2
accordingly.
1. Become root since all operations that follow require editing protected system files.
user@localhost:~$ sudo su Password:
root@localhost:~$
2. Edit /etc/security/limits.conf and add these two lines:

neo4j
neo4j
soft
hard
nofile
nofile
40000
40000
3. Edit /etc/pam.d/su and uncomment or add the following line:

session
required
pam_limits.so
4. A restart is required for the settings to take effect.

After the above procedure, the neo4j user will have a limit of 40000 simultaneous open files. If you
continue experiencing exceptions on Too many open files or Could not stat() directory then you
may have to raise that limit further.
406
Chapter23.High Availability
Note
The High Availability features are only available in the Neo4j Enterprise Edition.
Neo4j High Availability or Neo4j HA provides the following two main features:
1. It enables a fault-tolerant database architecture, where several Neo4j slave databases can be
configured to be exact replicas of a single Neo4j master database. This allows the end-user system
to be fully functional and both read and write to the database in the event of hardware failure.
2. It enables a horizontally scaling read-mostly architecture that enables the system to handle more
read load than a single Neo4j database instance can handle.
407
High Availability
23.1.Architecture
Neo4j HA has been designed to make the transition from single machine to multi machine operation
simple, by not having to change the already existing application.
Consider an existing application with Neo4j embedded and running on a single machine. To deploy
such an application in a multi machine setup the only required change is to switch the creation of the
GraphDatabaseService from GraphDatabaseFactory to HighlyAvailableGraphDatabaseFactory. Since both
implement the same interface, no additional changes are required.
Figure23.1.Multiple Neo4j instances in HA mode
When running Neo4j in HA mode there is always a single master and zero or more slaves. Compared
to other master-slave replication setups Neo4j HA can handle write requests on all machines so there
is no need to redirect those to the master specifically.
408
High Availability
A slave will handle writes by synchronizing with the master to preserve consistency. Writes to master
can be configured to be optimistically pushed to 0 or more slaves. By optimistically we mean the
master will try to push to slaves before the transaction completes but if it fails the transaction will still
be successful (different from normal replication factor). All updates will however propagate from the
master to other slaves eventually so a write from one slave may not be immediately visible on all other
slaves. This is the only difference between multiple machines running in HA mode compared to single
machine operation. All other ACID characteristics are the same.
409
High Availability
23.2.Setup and configuration

Neo4j HA can be set up to accommodate differing requirements for load, fault tolerance and available
hardware.
In HA mode, Neo4j instances form a cluster. The instances monitor each others' availability to take
account of instances joining and leaving the cluster. They elect one instance to be the master, and
designate the other instances to be slaves.
For installation instructions of a High Availability cluster see Section23.6, High Availability setup
tutorial.
23.2.1.Specifying cluster members

Specify the instances that should form the cluster by supplying ha.initial_hosts, a comma-separated
list of URLs. When each instance starts, if it can contact any of the initial hosts, then it will form a
cluster with them, otherwise it will start its own cluster.
Note that the parameter is called ha.initial_hosts because its only used when instances initially join
the cluster. This means that you can extend the cluster without changing the configuration of existing
instances.
23.2.2.Server configuration
If you are running Neo4j server, specify org.neo4j.server.database.mode=HA in neo4j-server.properties.
HA server configuration parameters
Parameter Name
Description
Example value
Required?
org. neo4j. server. database. mode
Whether to run as a
single server or in HA
mode.
single or HA
yes
23.2.3.Database configuration
HA configuration parameters should be supplied alongside general Neo4j parameters in
neo4j.properties. There are many configurable parameters, most in most cases it isnt necessary
to modify the default values. The only parameters that need to be specified are ha.server_id and
ha.initial_hosts.
HA database configuration parameters
Parameter Name
Description
Example value
Required?
ha. server_id
Id for a cluster
instance. Must be
unique within the
cluster.
yes
ha. initial_hosts
A comma-separated list server1:5001,

of other members of
server2:5001
the cluster to join.
yes
ha. cluster_server
Host & port to bind the server1:5001

cluster management
communication.
no
410
High Availability
Parameter Name
Description
Example value
ha. allow_init_cluster
Whether to allow this true

instance to create a
cluster if unable to join.
no
ha. default_timeout
Default timeout used

5s
for clustering timeouts.
Override specific
timeout settings
with proper values
if necessary. This
value is the default
value for settings
ha.heartbeat_interval,
ha.paxos_timeout and
ha.learn_timeout.
no
ha. heartbeat_interval
How often heartbeat

messages should
be sent. Defaults to
ha.default_timeout.
no
ha. heartbeat_timeout
Timeout for heartbeats 11s

between cluster
members. Should be
at least twice that of
ha.heartbeat_interval.
no
ha. broadcast_timeout
Timeout for
30s
broadcasting values in
cluster. Must consider
end-to-end duration of
Paxos algorithm. This
value is the default
value for settings
ha.join_timeout and
ha.leave_timeout.
no
ha. join_timeout
Timeout for joining

a cluster. Defaults to
ha.broadcast_timeout.
no
ha. configuration_timeout
Timeout for waiting

1s
for configuration from
an existing cluster
member during cluster
join.
no
ha. leave_timeout
Timeout for waiting

for cluster leave to
finish. Defaults to
ha.broadcast_timeout.
no
411
5s
30s
30s
Required?
High Availability
Parameter Name
Description
Example value
ha. paxos_timeout
Default timeout
5s
for all Paxos
timeouts. Defaults to
ha.default_timeout.
This value is the
default value
for settings
ha.phase1_timeout,
ha.phase2_timeout and
ha.election_timeout.
no
ha. phase1_timeout
Timeout for Paxos

phase 1. Defaults to
ha.paxos_timeout.
5s
no
ha. phase2_timeout
Timeout for Paxos

phase 2. Defaults to
ha.paxos_timeout.
5s
no
ha. learn_timeout
Timeout for learning

values. Defaults to
ha.default_timeout.
5s
no
ha. election_timeout
Timeout for waiting

for other members
to finish a role
election. Defaults to
ha.paxos_timeout.
5s
no
ha. read_timeout
How long a slave will 20s

wait for response from
master before giving
up.
no
ha. state_switch_timeout
Timeout for waiting

20s
for instance to become
master or slave.
no
ha. lock_read_timeout
Timeout for taking

remote (write) locks
on slaves. Defaults to
ha.read_timeout.
no
20s
ha.
Maximum number of 20
max_concurrent_channels_per_slave
connections a slave can
Required?
no
have to the master.
ha. server
Hostname and port to

bind the HA server.
ha. slave_only
Whether this instance false

should only participate
as slave in cluster. If
412
my-domain. com:6001
no
no
High Availability
Parameter Name
Description
Example value
set to true, it will never
be elected as master.
Required?
ha. branched_data_policy
Policy for how to

handle branched data.
no
ha. com_chunk_size
Max size of the data

2M
chunks that flows
between master
and slaves in HA.
Bigger size may
increase throughput,
but may be more
sensitive to variations
in bandwidth,
whereas lower size
increases tolerance for
bandwidth variations.
no
ha. pull_interval
Interval of pulling
updates from master.
10s
no
ha. tx_push_factor
The amount of slaves

the master will ask to
replicate a committed
transaction.
no
ha. tx_push_strategy
Push strategy of a
transaction to a slave
during commit.
fixed or round_robin
no
413
keep_last or keep_all
or keep_none
High Availability
23.3.How Neo4j HA operates

A Neo4j HA cluster operates cooperatively each database instance contains the logic needed in
order to coordinate with the other members of the cluster. On startup a Neo4j HA database instance
will try to connect to an existing cluster specified by configuration. If the cluster exists, the instance
will join it as a slave. Otherwise the cluster will be created and the instance will become its master.
When performing a write transaction on a slave each write operation will be synchronized with the
master (locks will be acquired on both master and slave). When the transaction commits it will first be
committed on the master and then, if successful, on the slave. To ensure consistency, a slave has to be
up to date with the master before performing a write operation. This is built into the communication
protocol between the slave and master, so that updates will be applied to a slave communicating with
its master automatically.
Write transactions performed directly through the master will execute in the same way as running
in normal non-HA mode. On success the transaction will be pushed out to a configurable number of
slaves (default one slave). This is done optimistically meaning if the push fails the transaction will
still be successful. Its also possible to configure push factor to 0 for higher write performance when
writing directly through the master, although increasing the risk of losing any transaction not yet
pulled by another slave if the master goes down.
Slaves can also be configured to pull updates asynchronously by setting the ha.pull_interval option.
Whenever a Neo4j database becomes unavailable, by means of for example hardware failure or
network outages, the other database instances in the cluster will detect that and mark it as temporarily
failed. A database instance that becomes available after being unavailable will automatically catch
up with the cluster. If the master goes down another (best suited) member will be elected and have
its role switched from slave to master after a quorum has been reached within the cluster. When the
new master has performed its role switch it will broadcast its availability to all the other members of
the cluster. Normally a new master is elected and started within just a few seconds and during this
time no writes can take place (the writes will block or in rare cases throw an exception). The only
time this is not true is when an old master had changes that did not get replicated to any other member
before becoming unavailable. If the new master is elected and performs changes before the old master
recovers, there will be two "branches" of the database after the point where the old master became
unavailable. The old master will move away its database (its "branch") and download a full copy from
the new master, to become available as a slave in the cluster.
All this can be summarized as:
Write transactions can be performed on any database instance in a cluster.
Neo4j HA is fault tolerant and can continue to operate from any number of machines down to a
single machine.
Slaves will be automatically synchronized with the master on write operations.
If the master fails a new master will be elected automatically.
The cluster automatically handles instances becoming unavailable (for example due to network
issues), and also makes sure to accept them as members in the cluster when they are available
again.
Transactions are atomic, consistent and durable but eventually propagated out to other slaves.
Updates to slaves are eventual consistent by nature but can be configured to be pushed
optimistically from master during commit.
If the master goes down any running write transaction will be rolled back and new transactions will
block or fail until a new master has become available.
414
High Availability
Reads are highly available and the ability to handle read load scales with more database instances in
the cluster.
415
High Availability
23.4.Arbiter Instances
A typical deployment of Neo4j will use a cluster of 3 machines to provide fault-tolerance and read
scalability. This setup is described in Section23.6, High Availability setup tutorial.
While having at least 3 instances is necessary for failover to happen in case the master becomes
unavailable, it is not required for all instances to run the full Neo4j stack, which includes the database
engine. Instead, what is called arbiter instances can be deployed. They can be regarded as cluster
participants in that their role is to take part in master elections with the single purpose of breaking ties
in the election process. That makes possible a scenario where you have a cluster of 2 Neo4j database
instances and an additional arbiter instance and still enjoy tolerance of a single failure of either of the
3 instances.
Arbiter instances are configured in the same way as Neo4j HA members are through the
neo4j.properties file in the installations conf/ directory. Settings that are not cluster specific are of
course ignored, so you can easily start up an arbiter instance in place of a properly configured Neo4j
instance.
To start an arbiter instance, call
neo4j_home$ ./bin/neo4j-arbiter start
You can also stop, install and remove it as a service and ask for its status in exactly the same way as
for Neo4j instances. See also Section21.2, Server Installation.
416
High Availability
23.5.Upgrading a Neo4j HA Cluster

This document describes the steps required to upgrade a Neo4j cluster without disrupting its
operation. This process is referred to as a rolling upgrade.
23.5.1.Upgrading from 1.9.x to 2.0

It is currently not possible to do a rolling upgrade from 1.9.x to 2.0. This means that in order for you
to upgrade a 1.9.x cluster to 2.0, you will incur some downtime.
Steps
In order to upgrade a cluster from 1.9.x to 2.0, you will need to do the following manually:
1. Backup your data!
2. Shut down every cluster member except the master.
3. Install the new version of Neo4j for each slave instance removing the data directory, but keeping
the configuration.
4. Shut down the master this brings your cluster completely offline!
5. Follow the instructions found in Section21.4.2, Explicit Upgrade for just the master
instance the data in master instance have now been converted to the new Neo4j 2.0 format.
6. Shut down the master instance again.
7. Distribute the data directory from master to each slave instance this saves time by not having to
re-build stores and indexes for each slave.
8. Bring master online.
9. Bring slaves online.
Your entire cluster has now been completely converted to Neo4j 2.0.
23.5.2.Upgrading from 1.8.x to 1.9

The starting assumptions are that there exists a cluster running Neo4j version 1.8 or newer with the
corresponding ZooKeeper instances and that the machine which is currently the master is known. It
is also assumed that on each machine the Neo4j service and the neo4j coordinator service is installed
under a directory which from here on is assumed to be /opt/old-neo4j
The process consists of upgrading each machine in turn by removing it from the cluster, moving over
the database and starting it back up again. Configuration settings also have to be transferred. It is
important to note that the last machine to be upgraded must be the master. In general, the "cluster
version" is defined by the version of the master, providing the master is of the older version the cluster
as a whole can operate (the 1.9 instances running in compatibility mode). When a 1.9 instance is
elected master however, the older instances are not capable of communicating with it, so we have
to make sure that the last machine upgraded is the old master. The upgrade process is detected
automatically from the joining 1.9 instances and they will not participate in a master election while
even a single old instance is part of the cluster.
Step 1: On each slave perform the upgrade
Download and unpack the new version. Copy over any configuration settings you run your
instances with, taking care for deprecated settings and API changes that can occur between versions.
Also, ensure that newly introduced settings have proper values (see Section23.2, Setup and
configuration). The most important thing about the settings setup is the ha.coordinators setting in
neo4j.properties which must be set to the value the existing 1.8 instances are using. You also have
417
High Availability
to make sure that all but one instance have the ha.allow_init_cluster setting to false - the machine
that has it set to true should be the one that is to become the new master. In addition, it is necessary
that the last machine to be upgraded (the 1.8 master) does not have the ha.coordinators setting present
in its configuration file. Finally, dont forget to copy over any server plugins you may have. First,
shutdown the neo4j instance with
service neo4j-service stop
Next, uninstall it
service neo4j-service remove
Now you can copy over the database. Assuming the old instance is at /opt/old-neo4j and the newly
unpacked under /opt/neo4j-enterprise-1.9 the proper command would be
cp -R /opt/old-neo4j/data/graph.db /opt/neo4j-enterprise-1.9/data/
Next install the neo4j service, which also starts it

/opt/neo4j-enterprise-1.9/bin/neo4j install
Done. Now check that the services are running and that webadmin reports the version 1.9.
Transactions should also be applied from the master as usual.
Step 2: Upgrade the master, complete the procedure
Warning
Make sure that the installation that will replace the current master instance does not have
ha.coordinators setting present in the neo4j.properties file.
Go to the current master and execute step 1 The moment it will be stopped another instance will take
over (the one with the allow_init_cluster setting set to true), transitioning the cluster to 1.9. Finish
Step 1 on this machine as well and you will have completed the process.
Step 3: Cleanup, removing the coordinator services
Each 1.8 installation still has a coordinator service installed and running. To have those removed you
need to execute at every upgraded instance
service neo4j-coordinator stop
service neo4j-coordinator remove
After that, the 1.8 instances are no longer active or needed and can be removed or archived.
23.5.3.Upgrading within the 1.9.x series

Warning
Due to an bug in the HA code, it may not be possible to do a rolling (i.e. uninterrupted)
upgrade to Neo4j 1.9.2 on 1.9 or 1.9.1 clusters. Attempting to do so may lead to an
unstable cluster and data loss may occur. It is suggested that upgrades to 1.9.2 happen
offline, where all instances are shutdown, upgraded and restarted. Upgrading from 1.9.2 to
any other version works as described elsewhere in this guide.
Upgrading between 1.9.x versions follows the same general pattern as described in the first part of this
guide, but is much simpler because of the compatibility of the configuration options between 1.9.x
releases. We will describe a step by step procedure, aimed at reducing the master switches to a single
change.
418
High Availability
Step 1: On each slave perform the upgrade
Download and unpack the new version. Copy over any configuration settings, ensuring that newly
introduced settings have proper values (see Section23.2, Setup and configuration). Dont forget to
copy over any server plugins you may have.
First, shutdown the neo4j instance with
service neo4j-service stop
Next, uninstall it
service neo4j-service remove
Now you can copy over the database. Assuming the old instance is at /opt/old-neo4j and the newly
unpacked under /opt/neo4j-enterprise-1.9.x the proper command would be
cp -R /opt/old-neo4j/data/graph.db /opt/neo4j-enterprise-1.9.x/data/
Next install the neo4j service, which also starts it

/opt/neo4j-enterprise-1.9.x/bin/neo4j install
Now check that the services are running and that webadmin reports the version 1.9.x. Transactions
should also be applied from the master as usual.
Step 2: Upgrade the master, complete the procedure
Go to the current master and execute step 1 The moment it will be stopped another instance will take
over, transitioning the cluster to the new 1.9.x version. Finish Step 1 on this machine as well and you
will have completed the process.
419
High Availability
23.6.High Availability setup tutorial
This guide will help you understand how to configure and deploy a Neo4j High Availability cluster.
Two scenarios will be considered:
Configuring 3 instances to be deployed on 3 separate machines, in a setting similar to what might
be encountered in a production environment.
Modifying the former to make it possible to run a cluster of 3 instances on the same physical
machine, which is particularly useful during development.
23.6.1.Background
Each instance in a Neo4j HA cluster must be assigned an integer ID, which serves as its unique
identifier. At startup, a Neo4j instance contacts the other instances specified in the ha.initial_hosts
configuration option.
When an instance establishes a connection to any other, it determines the current state of the cluster
and ensures that it is eligible to join. To be eligible the Neo4j instance must host the same database
store as other members of the cluster (although it is allowed to be in an older state), or be a new
deployment without a database store.
Explicitly configure IP Addresses/Hostnames for a cluster
Neo4j will attempt to configure IP addresses for itself in the absence of explicit
configuration. However in typical operational environments where machines have multiple
network cards and support IPv4 and IPv6 it is strongly recommended that the operator
explicitly sets the IP address/hostname configuration for each machine in the cluster.
Lets examine the available settings and the values they accept.
ha.server_id
is the cluster identifier for each instance. It must be a positive integer and must be unique
among all Neo4j instances in the cluster.
ha.server_id
For example, ha.server_id=1.

ha.cluster_server
is an address/port setting that specifies where the Neo4j instance will listen for
cluster communications (like hearbeat messages). The default port is 5001. In the absence of a
specified IP address, Neo4j will attempt to find a valid interface for binding. While this behavior
typically results in a well-behaved server, it is strongly recommended that users explicitly choose an
IP address bound to the network interface of their choosing to ensure a coherent cluster deployment.
ha.cluster_server
For example, ha.cluster_server=192.168.33.22:5001 will listen for cluster communications on the

network interface bound to the 192.168.33.0 subnet on port 5001.
ha.initial_hosts
is a comma separated list of address/port pairs, which specify how to reach other
Neo4j instances in the cluster (as configured via their ha.cluster_server option). These hostname/ports
will be used when the Neo4j instances starts, to allow it up to find and join the cluster. Specifying an
instances own address is permitted.
ha.initial_hosts
Warning
Do not use any whitespace in this configuration option.

420
High Availability
For example, ha.initial_hosts=192.168.33.22:5001,192.168.33.21:5001 will attempt to reach
Neo4j instances listening on 192.168.33.22 on port 5001 and 192.168.33.21 on port 5001 on the
192.168.33.0 subnet.
ha.server
is an address/port setting that specifies where the Neo4j instance will listen for transactions
(changes to the graph data) from the cluster master. The default port is 6001. In the absence of a
specified IP address, Neo4j will attempt to find a valid interface for binding. While this behavior
typically results in a well-behaved server, it is strongly recommended that users explicitly choose an
IP address bound to the network interface of their choosing to ensure a coherent cluster topology.
ha.server
ha.server
must user a different port to ha.cluster_server.
For example, ha.server=192.168.33.22:6001 will listen for cluster communications on the network
interface bound to the 192.168.33.0 subnet on port 6001.
Address/port format
The ha.cluster_server and ha.server configuration options are specified as <IP

address>:<port>.
For ha.server the IP address MUST be the address assigned to one of the hosts network
interfaces.
For ha.cluster_server the IP address MUST be the address assigned to one of the hosts
network interfaces, or the value 0.0.0.0, which will cause Neo4j to listen on every network
interface.
Either the address or the port can be omitted, in which case the default for that part will be
used. If the address is omitted, then the port must be preceded with a colon (eg. :5001).
The syntax for setting the port range is: <hostname>:<first port>[-<second port>]. In this
case, Neo4j will test each port in sequence, and select the first that is unused. Note that
this usage is not permitted when the hostname is specified as 0.0.0.0 (the "all interfaces"
address).
23.6.2.Getting started: Setting up a production cluster

Download and unpack Neo4j Enterprise
Download Neo4j Enterprise from the Neo4j download site <http://neo4j.org/download>, and unpack
on 3 separate machines.
Configure HA related settings for each installation
The following settings should be configured for each Neo4j installation. Note that all 3 installations
have the same configuration, except for the ha.server_id property.
Neo4j instance #1 neo4j-01.local
conf/neo4j.properties:
Unique server id for this Neo4j instance
can not be negative id and must be unique
ha.server_id = 1
List of other known instances in this cluster
ha.initial_hosts = neo4j-01.local:5001,neo4j-02.local:5001,neo4j-03.local:5001
Alternatively, use IP addresses:
ha.initial_hosts = 192.168.0.20:5001,192.168.0.21:5001,192.168.0.22:5001
421
High Availability
conf/neo4j-server.properties
Let the webserver only listen on the specified IP.
HA - High Availability
SINGLE - Single mode, default.
org.neo4j.server.database.mode=HA

ha.server_id = 2
ha.initial_hosts = 192.168.0.20:5001,192.168.0.21:5001,192.168.0.22:5001

ha.server_id = 3
ha.initial_hosts = 192.168.0.20:5001,192.168.0.21:5001,192.168.0.22:5001
Start the Neo4j Servers

Start the Neo4j servers as normal. Note the startup order does not matter.
neo4j-01$ ./bin/neo4j start
Startup Time
When running in HA mode, the startup script returns immediately instead of waiting for
the server to become available. This is because the instance does not accept any requests
422
High Availability
until a cluster has been formed. In the example above this happens when you startup
the second instance. To keep track of the startup state you can follow the messages in
console.log - the path to that is printed before the startup script returns.
Now, you should be able to access the 3 servers and check their HA status:
http://neo4j-01.local:7474/webadmin/#/info/org.neo4j/High%20Availability/
Tip
You can replace database #3 with an arbiter instance, see Arbiter Instances.
Thats it! You now have a Neo4j HA cluster of 3 instances running. You can start by making a
change on any instance and those changes will be propagated between them. For more HA related
configuration options take a look at HA Configuration.
23.6.3.Alternative setup: Creating a local cluster for testing

If you want to start a cluster similar to the one described above, but for development and testing
purposes, it is convenient to run all Neo4j instances on the same machine. This is easy to achieve,
although it requires some additional configuration as the defaults will conflict with each other.
Download and unpack Neo4j Enterprise
Download Neo4j Enterprise from the Neo4j download site <http://neo4j.org/download>, and unpack
into 3 separate directories on your test machine.
Configure HA related settings for each installation
The following settings should be configured for each Neo4j installation.
Neo4j instance #1 ~/neo4j-01
ha.server_id = 1
IP and port for this instance to bind to for communicating data with the
other neo4j instances in the cluster.
ha.server = 127.0.0.1:6363
online_backup_server = 127.0.0.1:6366
IP and port for this instance to bind to for communicating cluster information
with the other neo4j instances in the cluster.
ha.cluster_server = 127.0.0.1:5001
ha.initial_hosts = 127.0.0.1:5001,127.0.0.1:5002,127.0.0.1:5003
database location
http port (for all data, administrative, and UI access)
423
High Availability
https port (for all data, administrative, and UI access)

org.neo4j.server.webserver.https.port=7484

ha.server_id = 2
ha.server = 127.0.0.1:6364
ha.initial_hosts = 127.0.0.1:5001,127.0.0.1:5002,127.0.0.1:5003
database location

ha.server_id = 3
ha.server = 127.0.0.1:6365
ha.initial_hosts = 127.0.0.1:5001,127.0.0.1:5002,127.0.0.1:5003
database location
424
High Availability
Start the Neo4j Servers

Start the Neo4j servers as normal. Note the startup order does not matter.
localhost:~/neo4j-01$ ./bin/neo4j start
Now, you should be able to access the 3 servers and check their HA status:
http://127.0.0.1:7474/webadmin/#/info/org.neo4j/High%20Availability/
425
High Availability
23.7.REST endpoint for HA status information

23.7.1.Introduction
A common use case for Neo4j HA clusters is to direct all write requests to the master while using
slaves for read operations, distributing the read load across the cluster and and gain failover
capabilities for your deployment. The most common way to achieve this is to place a load balancer
in front of the HA cluster, an example being shown with HA Proxy. As you can see in that guide, it
makes use of a REST endpoint to discover which instance is the master and direct write load to it. In
this section, well deal with this REST endpoint and explain its semantics.
23.7.2.The endpoints
Each HA instance comes with 2 endpoints regarding its HA status. They are complimentary but each
may be used depending on your load balancing needs and your production setup. Those are:
/db/manage/server/ha/master
/db/manage/server/ha/slave
To use them, perform an HTTP GET operation on either and the following will be returned:
HA REST endpoint responses
Endpoint
/db/manage/server/ha/master
/db/manage/server/ha/slave
Instance
State
Returned Code
Body text
Master
200 OK
true
Slave
404 Not Found
false
Unknown
404 Not Found
UNKNOWN
Master
404 Not Found
false
Slave
200 OK
true
Unknown
404 Not Found
UNKNOWN
23.7.3.Examples
From the command line, a common way to ask those endpoints is to use curl. With no arguments, curl
will do an HTTP GET on the URI provided and will output the body text, if any. If you also want to get
the response code, just add the -v flag for verbose output. Here are some examples:
Requesting master endpoint on a running master with verbose output
> curl -v localhost:7474/db/manage/server/ha/master
* About to connect() to localhost port 7474 (0)
*
Trying ::1...
* connected
* Connected to localhost (::1) port 7474 (0)
> GET /db/manage/server/ha/master HTTP/1.1
> User-Agent: curl/7.24.0 (x86_64-apple-darwin12.0) libcurl/7.24.0 OpenSSL/0.9.8r zlib/1.2.5
> Host: localhost:7474
> Accept: */*
>
< HTTP/1.1 200 OK
< Content-Type: text/plain
< Access-Control-Allow-Origin: *
< Transfer-Encoding: chunked
426
High Availability
< Server: Jetty(6.1.25)
<
* Connection 0 to host localhost left intact
true* Closing connection 0
Requesting slave endpoint on a running master without verbose output:

> curl localhost:7474/db/manage/server/ha/slave
false
Finally, requesting the master endpoint on a slave with verbose output

> curl -v localhost:7475/db/manage/server/ha/master
* About to connect() to localhost port 7475 (0)
*
Trying ::1...
* connected
* Connected to localhost (::1) port 7475 (0)
> GET /db/manage/server/ha/master HTTP/1.1
> User-Agent: curl/7.24.0 (x86_64-apple-darwin12.0) libcurl/7.24.0 OpenSSL/0.9.8r zlib/1.2.5
> Host: localhost:7475
> Accept: */*
>
< HTTP/1.1 404 Not Found
< Content-Type: text/plain
< Access-Control-Allow-Origin: *
< Transfer-Encoding: chunked
< Server: Jetty(6.1.25)
<
* Connection 0 to host localhost left intact
false* Closing connection 0
Unknown status
The UNKNOWN status exists to describe when a Neo4j instance is neither master nor slave. For
example, the instance could be transitioning between states (master to slave in a recovery
scenario or slave being promoted to master in the event of failure). If the UNKNOWN status is
returned, the client should not treat the instance as a master or a slave and should instead
pick another instance in the cluster to use, wait for the instance to transit from the UNKNOWN
state, or undertake restorative action via systems admin.
427
High Availability
23.8.Setting up HAProxy as a load balancer

In the Neo4j HA architecture, the cluster is typically fronted by a load balancer. In this section we will
explore how to set up HAProxy to perform load balancing across the HA cluster.
For this tutorial we will assume a Linux environment with HAProxy already installed. See http://
haproxy.1wt.eu/ for downloads and installation instructions.
23.8.1.Configuring HAProxy
HAProxy can be configured in many ways. The full documentation is available at their website.
For this example, we will configure HAProxy to load balance requests to three HA servers. Simply
write the follow configuration to /etc/haproxy.cfg:
global
daemon
maxconn 256
defaults
mode http
timeout connect 5000ms
timeout client 50000ms
timeout server 50000ms
frontend http-in
bind *:80
default_backend neo4j
backend neo4j
server s1 10.0.1.10:7474 maxconn 32
listen admin
bind *:8080
stats enable
HAProxy can now be started by running:

/usr/sbin/haproxy -f /etc/haproxy.cfg
You can connect to http://<ha-proxy-ip>:8080/haproxy?stats to view the status dashboard. This

dashboard can be moved to run on port 80, and authentication can also be added. See the HAProxy
documentation for details on this.
23.8.2.Configuring separate sets for master and slaves

It is possible to set HAProxy backends up to only include slaves or the master. To accomplish this,
Neo4j provides health check URLs that HAProxy (or any load balancer for that matter) can use to
distinguish machines using HTTP response codes. The two URLs are:
/db/manage/server/ha/master, which returns 200 if the machine is the master, and 404 if the machine
is a slave
/db/manage/server/ha/slave, which returns 200 if the machine is a slave, and 404 if the machine is
the master
The following example excludes the master from the set of machines, so that requests are only be sent
to slaves.
global
428
High Availability
daemon
maxconn 256
defaults
mode http
frontend http-in
bind *:80
default_backend neo4j-slaves
backend neo4j-slaves
option httpchk GET /db/manage/server/ha/slave
server s1 10.0.1.10:7474 maxconn 32 check
listen admin
bind *:8080
stats enable
Note
In practice, writing to a slave is uncommon. While writing to slaves has the benefit of
ensuring that data is persisted in two places (the slave and the master), it comes at a
cost. The cost is that the slave must immediately become consistent with the master by
applying any missing transactions and then synchronously apply the new transaction with
the master. This is a more expensive operation than writing to the master and having the
master push changes to one or more slaves.
23.8.3.Cache-based sharding with HAProxy

Neo4j HA enables what is called cache-based sharding. If the dataset is too big to fit into the cache of
any single machine, then by applying a consistent routing algorithm to requests, the caches on each
machine will actually cache different parts of the graph. A typical routing key could be user ID.
In this example, the user ID is a query parameter in the URL being requested. This will route the same
user to the same machine for each request.
global
daemon
maxconn 256
defaults
mode http
frontend http-in
bind *:80
default_backend neo4j-slaves
backend neo4j-slaves
balance url_param user_id
listen admin
bind *:8080
429
High Availability
stats enable
Naturally the health check and query parameter-based routing can be combined to only route requests
to slaves by user ID. Other load balancing algorithms are also available, such as routing by source IP
(source), the URI (uri) or HTTP headers(hdr()).
430
Chapter24.Backup
Note
The Backup features are only available in the Neo4j Enterprise Edition.
431
Backup
24.1.Introducing Backup
Backups are performed over the network live from a running graph database onto a local copy. There
are two types of backup: full and incremental.
A full backup copies the database files without acquiring any locks, allowing for continued operations
on the target instance. This of course means that while copying, transactions will continue and the
store will change. For this reason, the transaction that was running when the backup operation started
is noted and, when the copy operation completes, all transactions from the latter down to the one
happening at the end of the copy are replayed on the backup files. This ensures that the backed up data
represent a consistent and up-to-date snapshot of the database storage.
In contrast, an incremental backup does not copy store files instead it copies the logs of the
transactions that have taken place since the last full or incremental backup which are then replayed
over an existing backup store. This makes incremental backups far more efficient than doing full
backups every time but they also require that a full backup has taken place before they are executed.
The backup tool will detect whether you are trying to run a full backup or an incremental one by
inspecting the target directory. Regardless of the mode a backup is created with, the resulting files
represent a consistent database snapshot and they can be used to boot up a Neo4j instance.
The database to be backed up is specified using a URI with syntax
<running mode>://<host>[:port]{,<host>[:port]*}
Running mode must be defined and is either single for non-HA or ha for HA clusters. The
<host>[:port] part points to a host running the database, on port port if not the default. The additional
host:port arguments are useful for passing multiple cluster members.
Important
As of version 1.9, backups are enabled by default. That means that the configuration
parameter online_backup_enabled defaults to true and that makes the backup service
available on the default port (6362). To enable the backup service on a different port use
online_backup_server=:9999.
432
Backup
24.2.Server and Embedded

To perform a backup from a running embedded or server database run:
Performing a full backup: create a blank directory and run the backup tool
mkdir /mnt/backup/neo4j-backup
./neo4j-backup -from single://192.168.1.34 -to /mnt/backup/neo4j-backup
Performing an incremental backup: just specify the location of your previous backup
./neo4j-backup -from single://192.168.1.34 -to /mnt/backup/neo4j-backup
Performing an incremental backup where the service is registered on a custom port
./neo4j-backup -from single://192.168.1.34:9999 -to /mnt/backup/neo4j-backup
433
Backup
24.3.Online Backup from Java

In order to programmatically backup your data full or subsequently incremental from a JVM based
program, you need to write Java code like
OnlineBackup backup = OnlineBackup.from( "127.0.0.1" );
backup.full( backupPath.getPath() );
backup.incremental( backupPath.getPath() );
For more information, please see the Javadocs for OnlineBackup <http://components.neo4j.org/neo4jenterprise/2.0.0/apidocs/org/neo4j/backup/OnlineBackup.html>
434
Backup
24.4.High Availability
To perform a backup on an HA cluster you specify one or more members of the target HA cluster.
Note that the addresses you must provide are the cluster server addresses and not the HA server
addresses. That is, use the value of the ha.cluster_server setting in the configuration.
Performing a backup from HA cluster, specifying only one cluster member
./neo4j-backup -from ha://192.168.1.15:5001 -to /mnt/backup/neo4j-backup
Performing a backup from HA cluster, specifying two possible cluster members
./neo4j-backup -from ha://192.168.1.15:5001,192.168.1.16:5002 -to /mnt/backup/neo4j-backup
435
Backup
24.5.Restoring Your Data

The Neo4j backups are fully functional databases. To use a backup, all you need to do replace your
database folder with the backup. Just make sure the database isnt running while replacing the folder.
436
Chapter25.Security
Neo4j in itself does not enforce security on the data level. However, there are different aspects that
should be considered when using Neo4j in different scenarios. See Section25.1, Securing access to
the Neo4j Server for details.
437
Security
25.1.Securing access to the Neo4j Server

25.1.1.Secure the port and remote client connection accepts
By default, the Neo4j Server is bundled with a Web server that binds to host localhost on port 7474,
answering only requests from the local machine.
This is configured in the conf/neo4j-server.properties file:
let the webserver only listen on the specified IP. Default
is localhost (only accept local connections). Uncomment to allow
any connection.
If you need to enable access from external hosts, configure the Web server in the conf/neo4jserver.properties by setting the property org.neo4j.server.webserver.address=0.0.0.0 to enable access
from any host.
25.1.2.Arbitrary code execution

By default, the Neo4j Server comes with some places where arbitrary code code execution can
happen. These are the Section19.15, Traversals REST endpoints. To secure these, either disable
them completely by removing offending plugins from the server classpath, or secure access to
these URLs through proxies or Authorization Rules. Also, the Java Security Manager, see http://
docs.oracle.com/javase/7/docs/technotes/guides/security/index.html can be used to secure parts of the
codebase.
25.1.3.HTTPS support
The Neo4j server includes built in support for SSL encrypted communication over HTTPS. The
first time the server starts, it automatically generates a self-signed SSL certificate and a private key.
Because the certificate is self signed, it is not safe to rely on for production use, instead, you should
provide your own key and certificate for the server to use.
To provide your own key and certificate, replace the generated key and certificate, or change the
neo4j-server.properties file to set the location of your certificate and key:
Certificate location (auto generated if the file does not exist)
org.neo4j.server.webserver.https.cert.location=ssl/snakeoil.cert
Private key location (auto generated if the file does not exist)
org.neo4j.server.webserver.https.key.location=ssl/snakeoil.key
Note that the key should be unencrypted. Make sure you set correct permissions on the private key, so
that only the Neo4j server user can read/write it.
Neo4j also supports chained SSL certificates. This requires to have all certificates in PEM format
combined in one file and the private key needs to be in DER format.
You can set what port the HTTPS connector should bind to in the same configuration file, as well as
turn HTTPS off:
Turn https-support on/off
org.neo4j.server.webserver.https.enabled=true
438
Security
25.1.4.Server Authorization Rules

Administrators may require more fine-grained security policies in addition to IP-level restrictions on
the Web server. Neo4j server supports administrators in allowing or disallowing access the specific
aspects of the database based on credentials that users or applications provide.
To facilitate domain-specific authorization policies in Neo4j Server, SecurityRules <http://
components.neo4j.org/neo4j-server/2.0.0/apidocs/org/neo4j/server/rest/security/SecurityRule.html>
can be implemented and registered with the server. This makes scenarios like user and role based
security and authentication against external lookup services possible.
Enforcing Server Authorization Rules
In this example, a (dummy) failing security rule is registered to deny access to all URIs to the server
by listing the rules class in neo4j-server.properties:
org.neo4j.server.rest.security_rules=my.rules.PermanentlyFailingSecurityRule
with the rule source code of:

public class PermanentlyFailingSecurityRule implements SecurityRule
{
public static final String REALM = "WallyWorld"; // as per RFC2617 :-)
@Override
public boolean isAuthorized( HttpServletRequest request )
{
return false; // always fails - a production implementation performs
// deployment-specific authorization logic here
}
@Override
public String forUriPath()
{
return "/*";
}
@Override
public String wwwAuthenticateHeader()
{
return SecurityFilter.basicAuthenticationResponse(REALM);
}
}
With this rule registered, any access to the server will be denied. In a production-quality
implementation the rule will likely lookup credentials/claims in a 3rd-party directory service (e.g.
LDAP) or in a local database of authorized users.
Example request
Example response
401: Unauthorized
WWW-Authenticate: Basic realm="WallyWorld"
439
Security
Using Wildcards to Target Security Rules
In this example, a security rule is registered to deny access to all URIs to the server by listing the
rule(s) class(es) in neo4j-server.properties. In this case, the rule is registered using a wildcard URI
path (where * characters can be used to signify any part of the path). For example /users* means the
rule will be bound to any resources under the /users root path. Similarly /users*type* will bind the
rule to resources matching URIs like /users/fred/type/premium.
org.neo4j.server.rest.security_rules=my.rules.PermanentlyFailingSecurityRuleWithWildcardPath

{
return "/protected/*";
}
With this rule registered, any access to URIs under /protected/ will be denied by the server. Using
wildcards allows flexible targeting of security rules to arbitrary parts of the servers API, including
any unmanaged extensions or managed plugins that have been registered.
Example request
GET http://localhost:7474/protected/tree/starts/here/dummy/more/stuff
Example response
401: Unauthorized
Using Complex Wildcards to Target Security Rules
In this example, a security rule is registered to deny access to all URIs matching a complex pattern.
The config looks like this:
org.neo4j.server.rest.security_rules=my.rules.PermanentlyFailingSecurityRuleWithComplexWildcardPath

public class PermanentlyFailingSecurityRuleWithComplexWildcardPath implements SecurityRule
{
public static final String REALM = "WallyWorld"; // as per RFC2617 :-)
@Override
public boolean isAuthorized( HttpServletRequest request )
{
return false;
}
@Override
{
return "/protected/*/something/else/*/final/bit";
}
@Override
public String wwwAuthenticateHeader()
{
return SecurityFilter.basicAuthenticationResponse(REALM);
440
Security
}
}
Example request
GET http://localhost:7474/protected/wildcard_replacement/x/y/z/something/else/
more_wildcard_replacement/a/b/c/final/bit/more/stuff
Example response
401: Unauthorized
25.1.5.Hosted Scripting
Important
The neo4j server exposes remote scripting functionality by default that allow full access to
the underlying system. Exposing your server without implementing a security layer poses a
substantial security vulnerability.
25.1.6.Security in Depth
Although the Neo4j server has a number of security features built-in (see the above chapters), for
sensitive deployments it is often sensible to front against the outside world it with a proxy like Apache
1
mod_proxy .
This provides a number of advantages:
Control access to the Neo4j server to specific IP addresses, URL patterns and IP ranges. This can be
used to make for instance only the /db/data namespace accessible to non-local clients, while the /
db/admin URLs only respond to a specific IP address.
<Proxy *>
Order Deny,Allow
Deny from all
Allow from 192.168.0
</Proxy>
While equivalent functionality can be implemented with Neo4js SecurityRule plugins (see above), for
operations professionals configuring servers like Apache is often preferable to developing plugins.
However it should be noted that where both approaches are used, they will work harmoniously
providing the behavior is consistent across proxy server and SecurityRule plugins.
Run Neo4j Server as a non-root user on a Linux/Unix system on a port < 1000 (e.g. port 80) using
ProxyPass /neo4jdb/data http://localhost:7474/db/data
ProxyPassReverse /neo4jdb/data http://localhost:7474/db/data
Simple load balancing in a clustered environment to load-balance read load using the Apache
2
mod_proxy_balancer plugin
<Proxy balancer://mycluster>
1
http://httpd.apache.org/docs/2.2/mod/mod_proxy.html
http://httpd.apache.org/docs/2.2/mod/mod_proxy_balancer.html
441
Security
BalancerMember http://192.168.1.50:80
BalancerMember http://192.168.1.51:80
</Proxy>
ProxyPass /test balancer://mycluster
442
Chapter26.Monitoring
Note
Most of the monitoring features are only available in the Enterprise edition of Neo4j.
In order to be able to continuously get an overview of the health of a Neo4j database, there are
different levels of monitoring facilities available. Most of these are exposed through JMX <http://
www.oracle.com/technetwork/java/javase/tech/javamanagement-140525.html>.
443
Monitoring
26.1.Adjusting remote JMX access to the Neo4j Server

Per default, the Neo4j Enterprise Server edition do not allow remote JMX connections, since the
relevant options in the conf/neo4j-wrapper.conf configuration file are commented out. To enable this
feature, you have to remove the characters from the various com.sun.management.jmxremote options
there.
When commented in, the default values are set up to allow remote JMX connections with certain
roles, refer to the conf/jmx.password, conf/jmx.access and conf/neo4j-wrapper.conf files for details.
Make sure that conf/jmx.password has the correct file permissions. The owner of the file has to be the
user that will run the service, and the permissions should be read only for that user. On Unix systems,
this is 0600.
On Windows, follow the tutorial at http://docs.oracle.com/javase/7/docs/technotes/guides/
management/security-windows.html to set the correct permissions. If you are running the service
under the Local System Account, the user that owns the file and has access to it should be SYSTEM.
With this setup, you should be able to connect to JMX monitoring of the Neo4j server using <IP-OFSERVER>:3637, with the username monitor and the password Neo4j.
Note that it is possible that you have to update the permissions and/or ownership of the conf/
jmx.password and conf/jmx.access files refer to the relevant section in conf/neo4j-wrapper.conf for
details.
Warning
For maximum security, please adjust at least the password settings in conf/jmx.password
for a production installation.
For more details, see: http://docs.oracle.com/javase/7/docs/technotes/guides/management/agent.html.
444
Monitoring
26.2.How to connect to a Neo4j instance using JMX and

JConsole
First, start your embedded database or the Neo4j Server, for instance using
$NEO4j_HOME/bin/neo4j start
Now, start JConsole with

$JAVA_HOME/bin/jconsole
Connect to the process running your Neo4j database instance:

Figure26.1.Connecting JConsole to the Neo4j Java process
Now, beside the MBeans exposed by the JVM, you will see an org.neo4j section in the MBeans tab.
Under that, you will have access to all the monitoring information exposed by Neo4j.
For opening JMX to remote monitoring access, please see Section26.1, Adjusting remote JMX
access to the Neo4j Server and the JMX documention <http://docs.oracle.com/javase/7/docs/
technotes/guides/management/agent.html>. When using Neo4j in embedded mode, make sure to pass
the com.sun.management.jmxremote.port=portNum or other configuration as JVM parameters to your
running Java process.
445
Monitoring
Figure26.2.Neo4j MBeans View
446
Monitoring
26.3.How to connect to the JMX monitoring

programmatically
In order to programmatically connect to the Neo4j JMX server, there are some convenience methods
in the Neo4j Management component to help you find out the most commonly used monitoring
attributes of Neo4j. See Section32.10, Reading a management attribute for an example.
Once you have access to this information, you can use it to for instance expose the values to SNMP
<http://en.wikipedia.org/wiki/Simple_Network_Management_Protocol> or other monitoring systems.
447
Monitoring
26.4.Reference of supported JMX MBeans

MBeans exposed by Neo4j
Branched Store: Information about the branched stores present in this HA cluster member
Cache/NodeCache: Information about the caching in Neo4j
Cache/RelationshipCache: Information about the caching in Neo4j
Configuration: The configuration parameters used to configure Neo4j
Diagnostics: Diagnostics provided by Neo4j
High Availability: Information about an instance participating in a HA cluster
Kernel: Information about the Neo4j kernel
Locking: Information about the Neo4j lock status
Memory Mapping: The status of Neo4j memory mapping
Primitive count: Estimates of the numbers of different kinds of Neo4j primitives
Store file sizes: Information about the sizes of the different parts of the Neo4j graph store
Transactions: Information about the Neo4j transaction manager
XA Resources: Information about the XA transaction manager
Note
For additional information on the primitive datatypes (int, long etc.) used in the JMX
attributes, please see Property value types in Section3.3, Properties.
MBean Branched Store (org.neo4j.management.BranchedStore) Attributes
Name
Description
Type
Read Write
Information about the branched stores present in this HA cluster member

BranchedStores
A list of the branched stores
org. neo4j.
management.
BranchedStoreInfo[]
as CompositeData[]
yes
no
MBean Cache/NodeCache (org.neo4j.management.Cache) Attributes

Name
Description
Type
Read Write
Information about the caching in Neo4j

CacheSize
The size of this cache (nr of entities or

total size in bytes)
long
yes
no
CacheType
The type of cache used by Neo4j
String
yes
no
HitCount
The number of times a cache query

returned a result
long
yes
no
MissCount
The number of times a cache query did

not return a result
long
yes
no
MBean Cache/NodeCache (org.neo4j.management.Cache) Operations

Name
Description
ReturnType
Signature
clear
Clears the Neo4j caches
void
(no parameters)
448
Monitoring
MBean Cache/RelationshipCache (org.neo4j.management.Cache) Attributes
Name
Description
Type
Read Write
Information about the caching in Neo4j

CacheSize
The size of this cache (nr of entities or

total size in bytes)
long
yes
no
CacheType
The type of cache used by Neo4j
String
yes
no
HitCount
The number of times a cache query

returned a result
long
yes
no
MissCount
The number of times a cache query did

not return a result
long
yes
no
MBean Cache/RelationshipCache (org.neo4j.management.Cache) Operations

Name
Description
ReturnType
Signature
clear
Clears the Neo4j caches
void
(no parameters)
MBean Configuration (org.neo4j.jmx.impl.ConfigurationBean) Attributes

Name
Description
Type
Read Write
The configuration parameters used to configure Neo4j

all_ stores_ total_
mapped_ memory_ size
The size to allocate for a memory

mapping pool to be shared between all
stores.
String
yes
no
allow_ store_ upgrade
Whether to allow a store upgrade in

String
case the current version of the database
starts against an older store version.
Setting this to true does not guarantee
successful upgrade, just that it allows an
attempt at it.
yes
no
array_ block_ size

String
arrays. This parameter is only honored
when the store is created, otherwise it is
ignored. The default block size is 120
bytes, and the overhead of each block
is the same as for string blocks, i.e., 8
bytes.
yes
no
backup_ slave
Mark this database as a backup slave.
String
yes
no
cache_ type
The type of cache to use for nodes

String
and relationships. Note that the Neo4j
Enterprise Edition has the additional hpc
cache type (High-Performance Cache).
See the chapter on caches in the manual
for more information.
yes
no
cypher_ parser_
version
Enable this to specify a parser other than String

the default one.
yes
no
449
Monitoring
Name
Description
Type
Read Write
dump_ configuration
Print out the effective Neo4j

configuration after startup.
String
yes
no
ephemeral
Configuration attribute
String
yes
no
forced_ kernel_ id
An identifier that uniquely identifies

this graph database instance within this
JVM. Defaults to an auto-generated
number depending on how many
instance are started in this JVM.
String
yes
no
gc_ monitor_ threshold
The amount of time in ms the monitor

String
thread has to be blocked before logging
a message it was blocked.
yes
no
gc_ monitor_ wait_ time
Amount of time in ms the GC monitor

thread will wait before taking another
measurement.
String
yes
no
ha. allow_ init_

cluster
Whether to allow this instance to create String

a cluster if unable to join.
yes
no
ha. branched_ data_

policy
Policy for how to handle branched data. String
yes
no
ha. broadcast_ timeout
Timeout for broadcasting values in

cluster. Must consider end-to-end
duration of Paxos algorithm. This
value is the default value for settings
ha.join_timeout and ha.leave_timeout.
String
yes
no
ha. cluster_ name
The name of a cluster.
String
yes
no
ha. cluster_ server
Host & port to bind the cluster

management communication.
String
yes
no
ha. com_ chunk_ size
Max size of the data chunks that flows

between master and slaves in HA.
Bigger size may increase throughput,
but may be more sensitive to variations
in bandwidth, whereas lower size
increases tolerance for bandwidth
variations.
String
yes
no
ha. configuration_
timeout
Timeout for waiting for configuration

String
from an existing cluster member during
cluster join.
yes
no
ha. default_ timeout
Default timeout used for clustering

String
timeouts. Override specific timeout
settings with proper values if necessary.
This value is the default value
for settings ha.heartbeat_interval,
ha.paxos_timeout and ha.learn_timeout.
yes
no
450
Monitoring
Name
Description
Type
Read Write
ha. election_ timeout
Timeout for waiting for other members

to finish a role election. Defaults to
ha.paxos_timeout.
String
yes
no
ha. heartbeat_
interval
How often heartbeat messages should be String

sent. Defaults to ha.default_timeout.
yes
no
ha. heartbeat_ timeout
Timeout for heartbeats between cluster

members. Should be at least twice that
of ha.heartbeat_interval.
String
yes
no
ha. initial_ hosts
A comma-separated list of other

members of the cluster to join.
String
yes
no
ha. join_ timeout
Timeout for joining a cluster. Defaults

to ha.broadcast_timeout.
String
yes
no
ha. learn_ timeout
Timeout for learning values. Defaults to String

ha.default_timeout.
yes
no
ha. leave_ timeout
Timeout for waiting for cluster leave to String

finish. Defaults to ha.broadcast_timeout.
yes
no
ha. lock_ read_ timeout
Timeout for taking remote (write) locks String

on slaves. Defaults to ha.read_timeout.
yes
no
ha. max_ concurrent_

channels_ per_ slave
Maximum number of connections a

slave can have to the master.
String
yes
no
ha. paxos_ timeout
Default timeout for all Paxos timeouts.

Defaults to ha.default_timeout. This
value is the default value for settings
ha.phase1_timeout, ha.phase2_timeout
and ha.election_timeout.
String
yes
no
ha. phase1_ timeout
Timeout for Paxos phase 1. Defaults to

ha.paxos_timeout.
String
yes
no
ha. phase2_ timeout
Timeout for Paxos phase 2. Defaults to

ha.paxos_timeout.
String
yes
no
ha. pull_ interval
Interval of pulling updates from master. String
yes
no
ha. read_ timeout
How long a slave will wait for response String

from master before giving up.
yes
no
ha. server_ id
Id for a cluster instance. Must be unique String

within the cluster.
yes
no
ha. server
Hostname and port to bind the HA

server.
String
yes
no
ha. slave_ only
Whether this instance should only

participate as slave in cluster. If set to
true, it will never be elected as master.
String
yes
no
ha. state_ switch_

timeout
Timeout for waiting for instance to

become master or slave.
String
yes
no
451
Monitoring
Name
Description
Type
ha. tx_ push_ factor
The amount of slaves the master will ask String

to replicate a committed transaction.
yes
no
ha. tx_ push_ strategy
Push strategy of a transaction to a slave String

during commit.
yes
no
intercept_ committing_
transactions
Determines whether any

TransactionInterceptors loaded will
intercept prepared transactions before
they reach the logical log.
String
yes
no
intercept_
deserialized_
transactions
Determines whether any

TransactionInterceptors loaded
will intercept externally received
transactions (e.g. in HA) before they
reach the logical log and are applied to
the store.
String
yes
no
jmx. port
Configuration attribute
String
yes
no
keep_ logical_ logs
Make Neo4j keep the logical transaction String

logs for being able to backup the
database.Can be used for specifying the
threshold to prune logical logs after. For
example "10 days" will prune logical
logs that only contains transactions
older than 10 days from the current
time, or "100k txs" will keep the 100k
latest transactions and prune any older
transactions.
yes
no
label_ block_ size

labels exceeding in-lined space in node
record. This parameter is only honored
when the store is created, otherwise it
is ignored. The default block size is 60
bytes, and the overhead of each block
is the same as for string blocks, i.e., 8
bytes.
String
yes
no
log_ mapped_ memory_

stats_ filename
The file where Neo4j will record

memory mapping statistics.
String
yes
no

stats_ interval
The number of records to be loaded

between regular logging of memory
mapping statistics.
String
yes
no

stats
Tell Neo4j to regularly log memory

mapping statistics.
String
yes
no
logging. threshold_
for_ rotation
Threshold in bytes for when database

logs (text logs, for debugging, that is)
are rotated.
String
yes
no
452
Read Write
Monitoring
Name
Description
Type
Read Write
logical_ log_
rotation_ threshold
Specifies at which file size the logical

log will auto-rotate. 0 means that no
rotation will automatically occur based
on file size. Default is 25M
String
yes
no
logical_ log
The base name for the logical log files, String

either an absolute path or relative to the
store_dir setting. This should generally
not be changed.
yes
no
mapped_ memory_ page_

size
Target size for pages of mapped

memory.
String
yes
no
neo_ store
The base name for the Neo4j Store files, String

either an absolute path or relative to the
store_dir setting. This should generally
not be changed.
yes
no
neostore. nodestore.
db. mapped_ memory
The size to allocate for memory

mapping the node store.
String
yes
no
neostore.
propertystore. db.
arrays. mapped_ memory

mapping the array property store.
String
yes
no
neostore.
propertystore. db.
index. keys. mapped_
memory

mapping the store for property key
strings.
String
yes
no
neostore.
propertystore. db.
index. mapped_ memory

mapping the store for property key
indexes.
String
yes
no
neostore.
propertystore. db.
mapped_ memory

mapping the property value store.
String
yes
no
neostore.
propertystore. db.
strings. mapped_
memory

mapping the string property store.
String
yes
no
neostore.
relationshipstore.
db. mapped_ memory

mapping the relationship store.
String
yes
no
node_ auto_ indexing
Controls the auto indexing feature

for nodes. Setting to false shuts
it down, while true enables it by
default for properties listed in the
node_keys_indexable setting.
String
yes
no
453
Monitoring
Name
Description
Type
Read Write
node_ keys_ indexable
A list of property names (comma

separated) that will be indexed by
default. This applies to Nodes only.
String
yes
no
online_ backup_
enabled
Enable support for running online

backups
String
yes
no
online_ backup_ server
Listening server for online backups
String
yes
no
read_ only
Only allow read operations from this

String
Neo4j instance. This mode still requires
write access to the directory for lock
purposes
yes
no
rebuild_ idgenerators_
fast
Use a quick approach for rebuilding

the ID generators. This give quicker
recovery time, but will limit the ability
to reuse the space of deleted entities.
String
yes
no
relationship_ auto_
indexing
Controls the auto indexing feature

for relationships. Setting to false
shuts it down, while true enables it
by default for properties listed in the
relationship_keys_indexable setting.
String
yes
no
relationship_ grab_
size
How many relationships to read at a

time during iteration
String
yes
no
relationship_ keys_
indexable
A list of property names (comma

separated) that will be indexed by
default. This applies to Relationships
only.
String
yes
no
remote_ logging_
enabled
Whether to enable logging to a remote

server or not.
String
yes
no
remote_ logging_ host
Host for remote logging using LogBack String

SocketAppender.
yes
no
remote_ logging_ port
Port for remote logging using LogBack

SocketAppender.
String
yes
no
remote_ shell_ enabled
Enable a remote shell server which shell String

clients can log in to
yes
no
store_ dir
The directory where the database files

are located.
String
yes
no
string_ block_ size

String
strings. This parameter is only honored
when the store is created, otherwise it
is ignored. Note that each character in a
string occupies two bytes, meaning that
a block size of 120 (the default size) will
hold a 60 character long string before
overflowing into a second block. Also
yes
no
454
Monitoring
Name
Description
Type
note that each block carries an overhead
of 8 bytes. This means that if the block
size is 120, the size of the stored records
will be 128 bytes.
Read Write
tx_ manager_ impl
The name of the Transaction Manager

service to use as defined in the TM
service provider constructor.
String
yes
no
use_ memory_ mapped_

buffers
Tell Neo4j to use memory mapped

buffers for accessing the native storage
layer.
String
yes
no
MBean Configuration (org.neo4j.jmx.impl.ConfigurationBean) Operations

Name
Description
ReturnType
Signature
apply
Apply settings
void
(no parameters)
MBean Diagnostics (org.neo4j.management.Diagnostics) Attributes

Name
Description
Type
Read Write
Diagnostics provided by Neo4j

DiagnosticsProviders
A list of the ids for the registered

diagnostics providers.
yes
List (java. util.

List)
no
MBean Diagnostics (org.neo4j.management.Diagnostics) Operations

Name
Description
ReturnType
Signature
dumpAll
Dump diagnostics information to JMX String
(no parameters)
dumpToLog
Dump diagnostics information to the

log.
void
(no parameters)
dumpToLog
Dump diagnostics information to the

log.
void
java. lang. String
extract
Operation exposed for management
String
java. lang. String
MBean High Availability (org.neo4j.management.HighAvailability) Attributes

Name
Description
Type
Read Write
Information about an instance participating in a HA cluster

Alive
Whether this instance is alive or not
boolean
yes
no
Available
Whether this instance is available or not boolean
yes
no
InstanceId
The identifier used to identify this server String

in the HA cluster
yes
no
InstancesInCluster
Information about all instances in this

cluster
yes
no
455
org. neo4j.
management.
ClusterMemberInfo[]
as CompositeData[]
Monitoring
Name
Description
Type
Read Write
LastCommittedTxId
The latest transaction id present in this

instances store
long
yes
no
LastUpdateTime
The time when the data on this instance String

was last updated from the master
yes
no
Role
The role this instance has in the cluster
yes
no
String
MBean High Availability (org.neo4j.management.HighAvailability) Operations

Name
Description
ReturnType
Signature
update
(If this is a slave) Update the database

on this instance with the latest
transactions from the master
String
(no parameters)
MBean Kernel (org.neo4j.jmx.Kernel) Attributes

Name
Description
Type
Read Write
Information about the Neo4j kernel

KernelStartTime
The time from which this Neo4j

instance was in operational mode.
Date (java. util.

Date)
yes
no
KernelVersion
The version of Neo4j
String
yes
no
MBeanQuery
An ObjectName that can be used as a

javax. management.
query for getting all management beans ObjectName
for this Neo4j instance.
yes
no
ReadOnly
Whether this is a read only instance
boolean
yes
no
StoreCreationDate
The time when this Neo4j graph store

was created.
Date (java. util.

Date)
yes
no
StoreDirectory
The location where the Neo4j store is

located
String
yes
no
StoreId
An identifier that, together with store

creation time, uniquely identifies this
Neo4j graph store.
String
yes
no
StoreLogVersion
The current version of the Neo4j store

logical log.
long
yes
no
MBean Locking (org.neo4j.management.LockManager) Attributes

Name
Description
Type
Read Write
Information about the Neo4j lock status

Locks
Information about all locks held by

Neo4j
NumberOf
AvertedDeadlocks
The number of lock sequences that

long
would have lead to a deadlock situation
456
java. util. List<org.

neo4j. kernel.
info. LockInfo> as
CompositeData[]
yes
no
yes
no
Monitoring
Name
Description
Type
that Neo4j has detected and averted (by
throwing DeadlockDetectedException).
Read Write
MBean Locking (org.neo4j.management.LockManager) Operations

Name
Description
ReturnType
Signature
getContendedLocks
getContendedLocks
java. util.
List<org. neo4j.
kernel. info.
LockInfo> as
CompositeData[]
long
MBean Memory Mapping (org.neo4j.management.MemoryMapping) Attributes

Name
Description
Type
Read Write
The status of Neo4j memory mapping

Get information about each pool of
org. neo4j.
memory mapped regions from store files management.
with memory mapping enabled
WindowPoolInfo[] as
MemoryPools
yes
no
CompositeData[]
MBean Primitive count (org.neo4j.jmx.Primitives) Attributes

Name
Description
Type
Read Write
Estimates of the numbers of different kinds of Neo4j primitives

NumberOf NodeIds InUse
An estimation of the number of nodes

used in this Neo4j instance
long
yes
no
NumberOf PropertyIds
InUse
An estimation of the number of

properties used in this Neo4j instance
long
yes
no
NumberOf
RelationshipIds InUse
An estimation of the number of

long
relationships used in this Neo4j instance
yes
no
NumberOf
RelationshipTypeIds
InUse
The number of relationship types used

in this Neo4j instance
long
yes
no
Type
Read Write
MBean Store file sizes (org.neo4j.jmx.StoreFile) Attributes

Name
Description
Information about the sizes of the different parts of the Neo4j graph store
ArrayStoreSize
The amount of disk space used to store

array properties, in bytes.
long
yes
no
LogicalLogSize
The amount of disk space used by the

current Neo4j logical log, in bytes.
long
yes
no
NodeStoreSize

nodes, in bytes.
long
yes
no
457
Monitoring
Name
Description
Type
Read Write
PropertyStoreSize

properties (excluding string values and
array values), in bytes.
long
yes
no
RelationshipStoreSize

relationships, in bytes.
long
yes
no
StringStoreSize

string properties, in bytes.
long
yes
no
TotalStoreSize
The total disk space used by this Neo4j

instance, in bytes.
long
yes
no
MBean Transactions (org.neo4j.management.TransactionManager) Attributes

Name
Description
Type
Read Write
Information about the Neo4j transaction manager

LastCommittedTxId
The id of the latest committed

transaction
long
yes
no
NumberOf Committed
Transactions
The total number of committed

transactions
long
yes
no
NumberOf Opened
Transactions
The total number started transactions
int
yes
no
NumberOf Open
Transactions
The number of currently open

transactions
int
yes
no
NumberOf RolledBack
Transactions
The total number of rolled back

transactions
long
yes
no
PeakNumberOf
Concurrent
Transactions
The highest number of transactions ever int

opened concurrently
yes
no
MBean XA Resources (org.neo4j.management.XaManager) Attributes

Name
Description
Type
Read Write
Information about the XA transaction manager

XaResources
Information about all XA resources

managed by the transaction manager
458
org. neo4j.
management.
XaResourceInfo[] as
CompositeData[]
yes
no
PartVI.Tools
The Tools part describes available Neo4j tools and how to use them.
Chapter27.Web Interface
The Neo4j Web Interface is the primary user interface for Neo4j.
The tool is available at http://127.0.0.1:7474/ after you have installed the Neo4j Server.
See the tool itself for more information!
460
Chapter28.Neo4j Shell
Neo4j shell is a command-line shell for running Cypher queries. Theres also commands to get
information about the database. In addition, you can browse the graph, much like how the Unix shell
along with commands like cd, ls and pwd can be used to browse your local file system.
Its a nice tool for development and debugging. This guide will show you how to get it going!
461
Neo4j Shell
28.1.Starting the shell
When used together with a Neo4j server, simply issue the following at the command line:
./bin/neo4j-shell
For the full list of options, see the reference in the Shell manual page.
To connect to a running Neo4j database, use Section28.1.4, Read-only mode for local databases
and see Section28.1.1, Enabling the shell server for remote databases.
You need to make sure that the shell jar file is on the classpath when you start up your Neo4j instance.
28.1.1.Enabling the shell server
Shell is enabled from the configuration of the Neo4j kernel, see Section22.2, Server Configuration.
Heres some sample configurations:
Enable the remote shell feature
remote_shell_enabled = true
The default port is 1337, but you can specify something else if you like
remote_shell_port = 1337
If you want to be a little protective of your data,
you can also tell the shell to only support read operations
remote_shell_read_only = true
When using the Neo4j server, see Section22.2, Server Configuration for how to add configuration
settings in that case.
There are two ways to start the shell, either by connecting to a remote shell server or by pointing it to
a Neo4j store path.
28.1.2.Connecting to a shell server
To start the shell and connect to a running server, run:

neo4j-shell
Alternatively supply -port and -name options depending on how the remote shell server was enabled.
Then youll get the shell prompt like this:
neo4j-sh (0)$
28.1.3.Pointing the shell to a path
To start the shell by just pointing it to a Neo4j store path you run the shell jar file. Given that the right
neo4j-kernel-<version>.jar and jta jar files are in the same path as your neo4j-shell-<version>.jar file
you run it with:
$ neo4j-shell -path path/to/neo4j-db
28.1.4.Read-only mode
By issuing the -readonly switch when starting the shell with a store path, changes cannot be made to
the database during the session.
$ neo4j-shell -readonly -path path/to/neo4j-db
28.1.5.Run a command and then exit
It is possible to tell the shell to just start, execute a command and then exit. This opens up for uses of
background jobs and also handling of huge output of f.ex. an ls command where you then could pipe
the output to less or another reader of your choice, or even to a file.
462
Neo4j Shell
And even to another neo4j-shell, e.g. for importing a dump of another database or cypher result. When
used with command mode the shell will not output a welcome message. So some examples of usage:
$ neo4j-shell -c "cd -a 24 && set name Mattias"
$ neo4j-shell -c "trav -r KNOWS" | less
28.1.6.Pass Neo4j configuration options

By setting the -config switch, you can provide a properties file that will be used to configure your
Neo4j instance, if started in embedded mode.
$ neo4j-shell -config conf/neo4j.properties -path mydb
28.1.7.Execute a file and then exit

To execute commands from a file and then exit just provide a -file filename. This is faster than
piping to the shell which still handles the input as if it was user input.
For example reading a dump file directly from the command line and executing it against the given
database. For example:
$ neo4j-shell -file export.cql > result.txt
Supplying - as the filename reads from stdin instead.
463
Neo4j Shell
28.2.Passing options and arguments

Passing options and arguments to your commands is very similar to many CLI commands in an *nix
environment. Options are prefixed with a - and can contain one or more options. Some options expect
a value to be associated with it. Arguments are string values which arent prefixed with -. Lets look at
ls as an example:
ls -r -f KNOWS:out -v 12345 will make a verbose listing of node 12345's outgoing relationships of type
KNOWS. The node id, 12345, is an argument to ls which tells it to do the listing on that node instead of
the current node (see pwd command). However a shorter version of this can be written:
ls -rfv KNOWS:out 12345. Here all three options are written together after a single - prefix. Even
though f is in the middle it gets associated with the KNOWS:out value. The reason for this is that the ls
command doesnt expect any values associated with the r or v options. So, it can infer the right values
for the rights options.
464
Neo4j Shell
28.3.Enum options
Some options expects a value which is one of the values in an enum, f.ex. direction part of
relationship type filtering where theres INCOMING, OUTGOING and BOTH. All such values can be supplied
in an easier way. Its enough that you write the start of the value and the interpreter will find what you
really meant. F.ex. out, in, i or even INCOMING.
465
Neo4j Shell
28.4.Filters
Some commands makes use of filters for varying purposes. F.ex. -f in ls and in trav. A filter is
supplied as a json <http://www.json.org/> object (w/ or w/o the surrounding {} brackets. Both keys
and values can contain regular expressions for a more flexible matching. An example of a filter could
be .*url.*:http.*neo4j.*,name:Neo4j. The filter option is also accompanied by the options -i and l which stands for ignore case (ignore casing of the characters) and loose matching (its considered
a match even if the filter value just matches a part of the compared value, not necessarily the entire
value). So for a case-insensitive, loose filter you can supply a filter with -f -i -l or -fil for short.
466
Neo4j Shell
28.5.Node titles
To make it easier to navigate your graph the shell can display a title for each node, f.ex. in ls -r. It
will display the relationships as well as the nodes on the other side of the relationships. The title is
displayed together with each node and its best suited property value from a list of property keys.
If youre standing on a node which has two KNOWS relationships to other nodes itd be difficult to know
which friend is which. The title feature addresses this by reading a list of property keys and grabbing
the first existing property value of those keys and displays it as a title for the node. So you may
specify a list (with or without regular expressions), f.ex: name,title.*,caption and the title for each
node will be the property value of the first existing key in that list. The list is defined by the client
(you) using the TITLE_KEYS environment variable (see Section28.6.7, Environment variables) and the
default being .*name.*,.*title.*
467
Neo4j Shell
28.6.How to use (individual commands)

The shell is modeled after Unix shells like bash that you use to walk around your local file system. It
has some of the same commands, like cd and ls. When you first start the shell (see instructions above),
you will get a list of all the available commands. Use man <command> to get more info about a particular
command. Some notes:
28.6.1.Comments
Single line comments, which will be ignored, can be made by using the prefix //. Example:
// This is a comment
28.6.2.Current node/relationship and path

You have a current node/relationship and a "current path" (like a current working directory in bash)
that youve traversed so far. When the shell first starts you are not positioned on any entity, but you
can cd your way through the graph (check your current path at any time with the pwd command). cd can
be used in different ways:
cd <node-id> will traverse one relationship to the supplied node id. The node must have a direct
relationship to the current node.
cd -a <node-id> will do an absolute path change, which means the supplied node doesnt have to
have a direct relationship to the current node.
cd -r <relationship-id> will traverse to a relationship instead of a node. The relationship must have
the current node as either start or end point. To see the relationship ids use the ls -vr command on
nodes.
cd -ar <relationship-id> will do an absolute path change which means the relationship can be any
relationship in the graph.
cd .. will traverse back one step to the previous location, removing the last path item from your
current path (pwd).
cd start (only if your current location is a relationship). Traverses to the start node of the
relationship.
cd end (only if your current location is a relationship). Traverses to the end node of the relationship.
28.6.3.Listing the contents of a node/relationship

List contents of the current node/relationship (or any other node) with the ls command. Please note
that it will give an empty output if the current node/relationship has no properties or relationships (for
example in the case of a brand new graph). ls can take a node id as argument as well as filters, see
Section28.4, Filters and for information about how to specify direction see Section28.3, Enum
options. Use man ls for more info.
28.6.4.Creating nodes and relationships

You create new nodes by connecting them with relationships to the current node. For example,
mkrel -t A_RELATIONSHIP_TYPE -d OUTGOING -c will create a new node (-c) and draw to it an OUTGOING
relationship of type A_RELATIONSHIP_TYPE from the current node. If you already have two nodes which
youd like to draw a relationship between (without creating a new node) you can do for example,
mkrel -t A_RELATIONSHIP_TYPE -d OUTGOING -n <other-node-id> and it will just create a new relationship
between the current node and that other node.
468
Neo4j Shell
28.6.5.Setting, renaming and removing properties

Property operations are done with the set, mv and rm commands. These commands operates on the
current node/relationship.
Use set <key> <value>, optionally with the -t option (for value type), to set a property. Supports every
type of value that Neo4j supports. Examples of a property of type int:
$ set -t int age 29
And an example of setting a double[] property:

$ set -t double[] my_values [1.4,12.2,13]
Example of setting a String property containing a JSON string:

mkrel -c -d i -t DOMAIN_OF --np "{'app':'foobar'}"
rm <key> removes a property.

mv <key> <new-key> renames a property from one key to another.
28.6.6.Deleting nodes and relationships

Deletion of nodes and relationships is done with the rmnode and rmrel commands. rmnode can delete
nodes, if the node to be deleted still has relationships they can also be deleted by supplying -f option.
rmrel can delete relationships, it tries to ensure connectedness in the graph, but relationships can be
deleted regardless with the -f option. rmrel can also delete the node on the other side of the deleted
relationship if its left with no more relationships, see -d option.
28.6.7.Environment variables
The shell uses environment variables a-la bash to keep session information, such as the current path
and more. The commands for this mimics the bash commands export and env. For example you can at
anytime issue a export STACKTRACES=true command to set the STACKTRACES environment variable to true.
This will then result in stacktraces being printed if an exception or error should occur. Allowed values
are all parseable JSON strings, so maps {age:10,name:"Mattias"} and arrays [1,2,3] are also supported.
Variables can also be assigned to each other. E.g. a=b will result in a containing the value of b.
This becomes especially interesting as all shell variables are automatically passed to cypher
statements as parameters. That makes it easy to query for certain start nodes or create nodes and
relationships with certain provided properties (as maps).
Values are removed by setting them to null or an empty value. List environment variables using env
28.6.8.Executing groovy/python scripts

The shell has support for executing scripts, such as Groovy <http://groovy.codehaus.org> and
Python <http://www.python.org> (via Jython <http://www.jython.org>). As of now the scripts
(*.groovy, *.py) must exist on the server side and gets called from a client with for example, gsh -renamePerson 1234 "Mathias" "Mattias" --doSomethingElse where the scripts renamePerson.groovy
and doSomethingElse.groovy must exist on the server side in any of the paths given by the GSH_PATH
environment variable (defaults to .:src:src/script). This variable is like the java classpath, separated by
a :. The python/jython scripts can be executed with the jsh in a similar fashion, however the scripts
have the .py extension and the environment variable for the paths is JSH_PATH.
When writing the scripts assume that theres made available an args variable (a String[]) which
contains the supplied arguments. In the case of the renamePerson example above the array would
469
Neo4j Shell
contain ["1234", "Mathias", "Mattias"]. Also please write your outputs to the out variable, such as
out.println( "My tracing text" ) so that it will be printed at the shell client instead of the server.
28.6.9.Traverse
You can traverse the graph with the trav command which allows for simple traversing from the
current node. You can supply which relationship types (w/ regex matching) and optionally direction
as well as property filters for matching nodes. In addition to that you can supply a command line
to execute for each match. An example: trav -o depth -r KNOWS:both,HAS_.*:incoming -c "ls $n".
Which means traverse depth first for relationships with type KNOWS disregarding direction and incoming
relationships with type matching HAS_.\* and do a ls <matching node> for each match. The node
filtering is supplied with the -f option, see Section28.4, Filters. See Section28.3, Enum options
for the traversal order option. Even relationship types/directions are supplied using the same format as
filters.
28.6.10.Query with Cypher

You can use Cypher to query the graph. For that, use the match or start command. You can also use
create statements to create nodes and relationships and use the cypher VERSION prefix to select a certain
cypher version.
Tip
Cypher queries need to be terminated by a semicolon ;.

Cypher commands are given all shell variables as parameters and the special self parameter for the
current node or relationship.
start n = node(0) return n; will give you a listing of the node with ID 0
cypher 1.9 start n = node(0) return n; will execute the query with Cypher version 1.9
START n = node({self}) MATCH (n)-[:KNOWS](friend) RETURN friend; will return the nodes connected
to the current node.
START n=node({me}) CREATE friend={props}, (me)-[r:KNOWS](friend); will create the friend and the
relationship according to the variables available.
28.6.11.Listing Indexes and Constraints

The schema command allows to list all existing indexes and constraints together with their current
status.
Note
This command does not list legacy indexes. For working with legacy indexes, please see
Section28.6.12, Legacy Indexing.
List all indexes and constraints:
schema
List indexes or constraints on :Person nodes for the property name:

schema -l :Person -p name
The schema command supports the following parameters:

-l :Label only list indexes or constraints for the given label :Label
470
Neo4j Shell
-p propertyKey only list indexes or constraints for the given property key propertyKey
-v if an index is in the FAILED state, print a verbose error cause if available
Indexes and constraints can be created or removed using Cypher or the Java Core API. They
are updated automatically whenever the graph is changed. See Section3.7, Schema for more
information.
28.6.12.Legacy Indexing
Its possible to query and manipulate legacy indexes via the index command.
Example: index -i persons name (will index the name for the current node or relationship in the
"persons" legacy index).
-g will do exact lookup in the legacy index and display hits. You can supply -c with a command to
be executed for each hit.
-q will ask the legacy index a query and display hits. You can supply -c with a command to be
executed for each hit.
--cd will change current location to the hit from the query. Its just a convenience for using the -c
option.
--ls will do a listing of the contents for each hit. Its just a convenience for using the -c option.
-i will index a key-value pair into a legacy index for the current node/relationship. If no value is
given the property value for that key for the current node is used as value.
-r will remove a key-value pair (if it exists) from a legacy index for the current node/relationship.
Key and value are optional.
-t will set the legacy index type to work with, for example index -t Relationship --delete friends
will delete the friends relationship index.
28.6.13.Transactions
It is useful to be able to test changes, and then being able to commit or rollback said changes.
Transactions can be nested. With a nested transaction, a commit does not write any changes to
disk, except for the top level transaction. A rollback, however works regardless of the level of the
transaction. It will roll back all open transactions.
begin transaction Starts a transaction.
commit Commits a transaction.
rollback Rollbacks all open transactions.
28.6.14.Dumping the database or a cypher result to Cypher statements

As a simple way of exporting a database or a subset of it, the dump command converts the graph of a
Cypher result or the whole database into a single Cypher create statement.
Note
Constraint export is not yet supported.

Examples:
dump dumps the whole database as single cypher create statement
471
Neo4j Shell
dump START n=node({self}) MATCH p=(n)-[r:KNOWS*](m) RETURN n,r,m; dumps the transitive
friendship graph of the current node.
neo4j-shell -path db1 -c dump START n=node:node_auto_index(name='Mattias\') MATCH p=(n)[r:KNOWS](m) RETURN p; | neo4j-shell -path db2 imports the subgraph of the first database (db1)
into the second (db2)
Example Dump Scripts
create a new node and go to it
neo4j-sh (?)$ mknode --cd --np "{'name':'Neo'}"
create a relationship
neo4j-sh (Neo,0)$ mkrel -c -d i -t LIKES --np "{'app':'foobar'}"
Export the cypher statement results
neo4j-sh (Neo,0)$ dump START n=node({self}) MATCH (n)-[r]-(m) return n,r,m;
begin
create (_0 {`name`:"Neo"})
create (_1 {àpp`:"foobar"})
create _1-[:`LIKES`]->_0
;
commit
create an index
neo4j-sh (?)$ create index on :Person(name);
+-------------------+
| No data returned. |
+-------------------+
Indexes added: 1
1120 ms
create one labeled node and a relationship

neo4j-sh (?)$ create (m:Person:Hacker {name:'Mattias'}), (m)-[:KNOWS]->(m);
+-------------------+
| No data returned. |
+-------------------+
Nodes created: 1
Properties set: 1
Labels added: 2
268 ms
Export the whole database including indexes

neo4j-sh (?)$ dump
begin
create index on :`Person`(`name`);
create (_0:`Person`:`Hacker` {`name`:"Mattias"})
create _0-[:`KNOWS`]->_0
;
commit
472
Neo4j Shell
28.7.An example shell session

Create a node
neo4j-sh (?)$ mknode --cd
where are we?
neo4j-sh (0)$ pwd
Current is (0)
(0)
On the current node, set the key "name" to value "Jon"

neo4j-sh (0)$ set name "Jon"
send a cypher query
neo4j-sh (Jon,0)$ start n=node(0) return n;
+---------------------+
| n
|
+---------------------+
| Node[0]{name:"Jon"} |
+---------------------+
1 row
377 ms
make an incoming relationship of type LIKES, create the end node with the node properties specified.
neo4j-sh (Jon,0)$ mkrel -c -d i -t LIKES --np "{'app':'foobar'}"
where are we?
neo4j-sh (Jon,0)$ ls
*name =[Jon]
(me)<-[:LIKES]-(1)
change to the newly created node

neo4j-sh (Jon,0)$ cd 1
list relationships, including relationship id
neo4j-sh (1)$ ls -avr
(me)-[:LIKES,0]->(Jon,0)
create one more KNOWS relationship and the end node

neo4j-sh (1)$ mkrel -c -d i -t KNOWS --np "{'name':'Bob'}"
print current history stack
neo4j-sh (1)$ pwd
Current is (1)
(Jon,0)-->(1)
verbose list relationships

neo4j-sh (1)$ ls -avr
(me)-[:LIKES,0]->(Jon,0)
(me)<-[:KNOWS,1]-(Bob,2)
473
Neo4j Shell
28.8.A Matrix example
This example is creating a graph of the characters in the Matrix via the shell and then executing
Cypher queries against it:
Figure28.1.Shell Matrix Example
ROOT
nam e = 'Thom as Andersson'
KNOWS
age = 3
KNOWS
nam e = 'Morpheus'
KNOWS
age = 90
nam e = 'Trinit y'
KNOWS
nam e = 'Cypher'
KNOWS
nam e = 'Agent Sm it h'
CODED_BY
nam e = 'The Archit ect '
Neo4j is configured for autoindexing, in this case with the following in the Neo4j configuration file:
node_auto_indexing=true
node_keys_indexable=name,age
relationship_auto_indexing=true
relationship_keys_indexable=ROOT,KNOWS,CODED_BY
The following is a sample shell session creating the Matrix graph and querying it.
Create a reference node
neo4j-sh (?)$ mknode --cd
474
Neo4j Shell
create the Thomas Andersson node
neo4j-sh (0)$ mkrel -t ROOT -c -v
Node (1) created
Relationship [:ROOT,0] created
go to the new node

neo4j-sh (0)$ cd 1
set the name property
neo4j-sh (1)$ set name "Thomas Andersson"
create Thomas direct friends
neo4j-sh (Thomas Andersson,1)$ mkrel -t KNOWS -cv
Node (2) created
Relationship [:KNOWS,1] created
go to the new node

neo4j-sh (Thomas Andersson,1)$ cd 2
neo4j-sh (2)$ set name "Trinity"
go back in the history stack
neo4j-sh (Trinity,2)$ cd ..
create Thomas direct friends
neo4j-sh (Thomas Andersson,1)$ mkrel -t KNOWS -cv
Node (3) created
Relationship [:KNOWS,2] created
go to the new node

neo4j-sh (Thomas Andersson,1)$ cd 3
neo4j-sh (3)$ set name "Morpheus"
create relationship to Trinity
neo4j-sh (Morpheus,3)$ mkrel -t KNOWS 2
list the relationships of node 3
neo4j-sh (Morpheus,3)$ ls -rv
(me)-[:KNOWS,3]->(Trinity,2)
(me)<-[:KNOWS,2]-(Thomas Andersson,1)
change the current position to relationship 2

neo4j-sh (Morpheus,3)$ cd -r 2
set the age property on the relationship
neo4j-sh [:KNOWS,2]$ set -t int age 3
back to Morpheus
neo4j-sh [:KNOWS,2]$ cd ..
next relationship
neo4j-sh (Morpheus,3)$ cd -r 3
set the age property on the relationship
neo4j-sh [:KNOWS,3]$ set -t int age 90
position to the start node of the current relationship
neo4j-sh [:KNOWS,3]$ cd start
475
Neo4j Shell
new node
neo4j-sh (Morpheus,3)$ mkrel -t KNOWS -c
list relationships on the current node
neo4j-sh (Morpheus,3)$ ls -r
(me)-[:KNOWS]->(Trinity,2)
(me)-[:KNOWS]->(4)
(me)<-[:KNOWS]-(Thomas Andersson,1)
go to Cypher
neo4j-sh (Morpheus,3)$ cd 4
set the name
neo4j-sh (4)$ set name Cypher
create new node from Cypher
neo4j-sh (Cypher,4)$ mkrel -ct KNOWS
list relationships
neo4j-sh (Cypher,4)$ ls -r
(me)-[:KNOWS]->(5)
(me)<-[:KNOWS]-(Morpheus,3)
go to the Agent Smith node

neo4j-sh (Cypher,4)$ cd 5
set the name
neo4j-sh (5)$ set name "Agent Smith"
outgoing relationship and new node
neo4j-sh (Agent Smith,5)$ mkrel -cvt CODED_BY
Node (6) created
Relationship [:CODED_BY,6] created
go there
neo4j-sh (Agent Smith,5)$ cd 6
set the name
neo4j-sh (6)$ set name "The Architect"
go to the first node in the history stack
neo4j-sh (The Architect,6)$ cd
Morpheus' friends, looking up Morpheus by name in the Neo4j autoindex
neo4j-sh (?)$ start morpheus = node:node_auto_index(name='Morpheus') match morpheus-[:KNOWS]-zionist return zionist.name;
+--------------------+
| zionist.name
|
+--------------------+
| "Trinity"
|
| "Cypher"
|
| "Thomas Andersson" |
+--------------------+
3 rows
322 ms
Morpheus' friends, looking up Morpheus by name in the Neo4j autoindex

neo4j-sh (?)$ cypher 2.0 start morpheus = node:node_auto_index(name='Morpheus') match morpheus-[:KNOWS]-zionist return zionist.name;
+--------------------+
| zionist.name
|
+--------------------+
476
Neo4j Shell
| "Trinity"
|
| "Cypher"
|
| "Thomas Andersson" |
+--------------------+
3 rows
289 ms
477
PartVII.Community
The Neo4j project has a strong community around it. Read about how to get help from the community
and how to contribute to it.
Chapter29.Community Support
You can learn a lot about Neo4j on different events. To get information on upcoming Neo4j events,
have a look here:
http://www.neo4j.org/
http://neo4j.meetup.com/
Get help from the Neo4j open source community; here are some starting points.
The neo4j tag at stackoverflow: http://stackoverflow.com/questions/tagged/neo4j

Neo4j Community Discussions: https://groups.google.com/forum/#!forum/neo4j
Twitter: https://twitter.com/neo4j
IRC channel: irc://irc.freenode.net/neo4j web chat <http://webchat.freenode.net/?
randomnick=1&channels=neo4j>.
Report a bug or add a feature request:

https://github.com/neo4j/neo4j/issues
Questions regarding the documentation: The Neo4j Manual is published online with a comment
function, please use that to post any questions or comments regarding the documentation. See http://
docs.neo4j.org/2.0.0.
479
Chapter30.Contributing to Neo4j
The Neo4j project is an Open Source effort to bring fast complex data storage and processing to life.
Every form of help is highly appreciated by the community - and you are not alone, see Section30.6,
Contributors!
One crucial aspect of contributing to the Neo4j project is the Section30.1, Contributor License
Agreement.
In short: make sure to sign the CLA and send in the email, or the Neo4j project wont be able to accept
your contribution.
Note that you can contribute to Neo4j also by contributing documentation or giving feedback on
the current documentation. Basically, at all the places where you can get help, theres also room for
contributions.
If you want to contribute, there are some good areas to start with, especially for getting in contact with
the community, Chapter29, Community Support.
To document your efforts, we highly recommend to read Section30.3, Writing Neo4j
Documentation.
480
Contributing to Neo4j
30.1.Contributor License Agreement

30.1.1.Summary
We require all source code that is hosted on the Neo4j infrastructure to be contributed through the
Neo4j Contributor License Agreement <http://dist.neo4j.org/neo4j-cla.pdf> (CLA). The purpose
of the Neo4j Contributor License Agreement is to protect the integrity of the code base, which
in turn protects the community around that code base: the founding entity Neo Technology, the
Neo4j developer community and the Neo4j users. This kind of contributor agreement is common
amongst free software and open source projects (it is in fact very similar to the widely signed Oracle
Contributor Agreement <http://www.oracle.com/technetwork/community/oca-486395.html>).
Please see the below or send a mail to admins [at] neofourjay.org if you have any other questions
about the intent of the CLA. If you have a legal question, please ask a lawyer.
30.1.2.Common questions
Am I losing the rights to my own code?
No, the Neo4j CLA <http://dist.neo4j.org/neo4j-cla.pdf> only asks you to share your rights, not
relinquish them. Unlike some contribution agreements that require you to transfer copyrights to
another organization, the CLA does not take away your rights to your contributed intellectual
property. When you agree to the CLA, you grant us joint ownership in copyright, and a patent license
for your contributions. You retain all rights, title, and interest in your contributions and may use them
for any purpose you wish. Other than revoking our rights, you can still do whatever you want with
your code.
What can you do with my contribution?
We may exercise all rights that a copyright holder has, as well as the rights you grant in the Neo4j
CLA <http://dist.neo4j.org/neo4j-cla.pdf> to use any patents you have in your contributions. As
the CLA provides for joint copyright ownership, you may exercise the same rights as we in your
contributions.
What are the community benefits of this?
Well, it allows us to sponsor the Neo4j projects and provide an infrastructure for the community,
while making sure that we can include this in software that we ship to our customers without any
nasty surprises. Without this ability, we as a small company would be hard pressed to release all our
code as free software.
Moreover, the CLA lets us protect community members (both developers and users) from hostile
intellectual property litigation should the need arise. This is in line with how other free software
stewards like the Free Software Foundation - FSF <http://www.fsf.org> defend projects (except
with the FSF, theres no shared copyright but instead you completely sign it over to the FSF). The
contributor agreement also includes a free software covenant, or a promise that a contribution will
remain available as free software.
At the end of the day, you still retain all rights to your contribution and we can stand confident that we
can protect the Neo4j community and the Neo Technology customers.
Can we discuss some items in the CLA?
Absolutely! Please give us feedback! But lets keep the legalese off the mailing lists. Please mail your
feedback directly to cla (@t) neotechnology dot cm and well get back to you.
481
I still dont like this CLA.
Thats fine. You can still host it anywhere else, of course. Please do! Were only talking here about the
rules for the infrastructure that we provide.
30.1.3.How to sign
When youve read through the CLA, please send a mail to cla (@t) neotechnology dot cm. Include
the following information:
Your full name.

Your e-mail address.
An attached copy of the Neo4j CLA <http://dist.neo4j.org/neo4j-cla.pdf>.
That you agree to its terms.
For example:
Hi. My name is John Doe (john@doe.com).
I agree to the terms in the attached Neo4j Contributor License Agreement.
482
30.2.Areas for contribution

Neo4j is a project with a vast ecosystem and a lot of space for contributions. Where you can and want
to pitch in depends of course on your time, skill set and interests. Below are some of the areas that
might interest you:
30.2.1.Neo4j Core Projects

The Neo4j open issues <https://github.com/neo4j/neo4j/issues> for some starting points for
contribution
See the GitHub Neo4j area <https://github.com/neo4j/> for a list of projects
30.2.2.Maintaining Neo4j Documentation

Some parts of the documentation need extra care from the community to stay up to date. They
typically refer to different kinds of community contributions. The easiest way to contribute fixes is to
comment at the online HTML version <http://docs.neo4j.org/chunked/snapshot/>.
30.2.3.Drivers and bindings to Neo4j

REST: see Chapter6, Languages for a list of active projects
483
30.3.Writing Neo4j Documentation

Note
Other than writing documentation, you can help out by providing comments - head over to
the online HTML version <http://docs.neo4j.org/chunked/snapshot/> to do that!
For how to build the manual see: readme <https://github.com/neo4j/neo4j/blob/master/manual/
README.asciidoc>
The documents use the asciidoc format, see:
Aciidoc Reference <http://www.methods.co.nz/asciidoc/>
AsciiDoc cheatsheet <http://powerman.name/doc/asciidoc>
The cheatsheet is really useful!
30.3.1.Overall Flow
Each (sub)project has its own documentation, which will produce a docs.jar file. By default this file is
assembled from the contents in src/docs/. Asciidoc documents have the .asciidoc file extension.
The documents can use code snippets which will extract code from the project. The corresponding
code must be deployed to the sources.jar or test-sources.jar file.
By setting up a unit test accordingly, documentation can be written directly in the JavaDoc comment.
The above files are all consumed by the build of the manual (by adding them as dependencies). To get
content included in the manual, it has to be explicitly included by a document in the manual as well.
Note that different ways to add documentation works best for different cases:
For detail level documentation, it works well to write the documentation as part of unit tests (in
the JavaDoc comment). In this case, you typically do not want to link to the source code in the
documentation.
For tutorial level documentation, the result will be best by writing a .asciidoc file containing
the text. Source snippets and output examples can then be included from there. In this case you
typically want to link to the source code, and users should be able to run it without any special
setup.
30.3.2.File Structure in docs.jar

Directory
Contents
dev/
content aimed at developers
dev/images/
images used by the dev docs
ops/
content aimed at operations
ops/images/
images used by the ops docs
man/
manpages
Additional subdirectories are used as needed to structure the documents, like dev/tutorial/, ops/
tutorial/ etc.
30.3.3.Headings and document structure
Each document starts over with headings from level zero (the document title). Each document should
have an id. In some cases sections in the document need to have ids as well, this depends on where
484
they fit in the overall structure. To be able to link to content, it has to have an id. Missing ids in
mandatory places will fail the build.
This is how a document should start:
[[unique-id-verbose-is-ok]]
= The Document Title =
To push the headings down to the right level in the output, the leveloffset attribute is used when
including the document inside of another document.
Subsequent headings in a document should use the following syntax:
== Subheading ==
... content here ...
=== Subsubheading ===
content here ...
Asciidoc comes with one more syntax for headings, but in this project its not used.
30.3.4.Writing
Put one sentence on each line. This makes it easy to move content around, and also easy to spot (too)
long sentences.
30.3.5.Gotchas
A chapter cant be empty. (the build will fail on the docbook xml validity check)
Always leave a blank line at the end of documents (or the title of the next document might end up in
the last paragraph of the document)
As {} are used for Asciidoc attributes, everything inside will be treated as an attribute. What you
have to do is to escape the opening brace: \{. If you dont, the braces and the text inside them will
be removed without any warning being issued!
30.3.6.Links
To link to other parts of the manual the id of the target is used. This is how such a reference looks:
<<community-docs-overall-flow>>
Which will render like: Section30.3.1, Overall Flow
Note
Just write "see <<target-id>>" and similar, that should suffice in most cases.
If you need to link to another document with your own link text, this is what to do:
<<target-id, link text that fits in the context>>
Note
Having lots of linked text may work well in a web context but is a pain in print, and we
aim for both!
External links are added like this:
http://neo4j.org/[Link text here]
Which renders like: Link text here <http://neo4j.org/>

485
For short links it may be better not to add a link text, just do:
http://neo4j.org/
Which renders like: http://neo4j.org/
Note
Its ok to have a dot right after the URL, it wont be part of the link.
30.3.7.Text Formatting
_Italics_ is rendered as Italics and used for emphasis.
*Bold* is rendered as Bold and used sparingly, for strong emphasis only.
+methodName()+ is rendered as methodName() and is used for literals as well (note: the content
between the + signs will be parsed).
`command` is rendered as command (typically used for command-line) (note: the content between the
` signs will not be parsed).
Mono++space++d is rendered as Monospaced and is used for monospaced letters.
'my/path/' is rendered as my/path/ (used for file names and paths).
``Double quoted'' (that is two grave accents to the left and two acute accents to the right) renders as
Double quoted.
`Single quoted' (that is a single grave accent to the left and a single acute accent to the right) renders
as Single quoted.
30.3.8.Admonitions
These are very useful and should be used where appropriate. Choose from the following (write all
caps and no, we cant easily add new ones):
Note
Note.
Tip
Tip.
Important
Important
Caution
Caution
Warning
Warning
Heres how its done:

NOTE: Note.
A multiline variation:
486
[TIP]
Tiptext.
Line 2.
Which is rendered as:
Tip
Tiptext. Line 2.
30.3.9.Images
Important
All images in the entire manual share the same namespace. You know how to handle that.
Images Files
To include an image file, make sure it resides in the images/ directory relative to the document youre
including it from. Then go:
image::neo4j-logo.png[]
Static Graphviz/DOT
We use the Graphviz/DOT language to describe graphs. For documentation see http://graphviz.org/.
This is how to include a simple example graph:
["dot", "community-docs-graphdb-rels.svg"]
---"Start node" -> "End node" [label="relationship"]
----

St art node
relat ionship
End node
Heres an example using some predefined variables available in the build:

["dot", "community-docs-graphdb-rels-overview.svg", "meta"]
---"A Relationship" [fillcolor="NODEHIGHLIGHT"]
"Start node" [fillcolor="NODE2HIGHLIGHT"]
"A Relationship" -> "Start node" [label="has a"]
"A Relationship" -> "End node" [label="has a"]
487
"A Relationship" -> "Relationship type" [label="has a"]
"Name" [TEXTNODE]
"Relationship type" -> "Name" [label="uniquely identified by" color="EDGEHIGHLIGHT" fontcolor="EDGEHIGHLIGHT"]
----

A Relat ionship
has a
St art node
has a
End node
has a
Relat ionship t ype
uniquely ident ified by

Nam e
The optional second argument given to the dot filter defines the style to use:
when not defined: Default styling for nodespace examples.
neoviz: Nodespace view generated by Neoviz.
meta: Graphs that dont resemble db contents, but rather concepts.
Caution
Keywords of the DOT language have to be surrounded by double quotes when used for
other purposes. The keywords include node, edge, graph, digraph, subgraph, and strict.
30.3.10.Attributes
Common attributes you can use in documents:
{neo4j-version} - rendered as "2.0.0"
{neo4j-git-tag} - rendered as "2.0.0"
{lucene-version} - rendered as "3_6_2"
These can substitute part of URLs that point to for example APIdocs or source code. Note that neo4jgit-tag also handles the case of snapshot/master.
Sample Asciidoc attributes which can be used:
{docdir} - root directory of the documents
{nbsp} - non-breaking space
30.3.11.Comments
Theres a separate build including comments. The comments show up with a yellow background. This
build doesnt run by default, but after a normal build, you can use make annotated to build it. You can
also use the resulting page to search for content, as the full manual is on a single page.
Heres how to write a comment:
// this is a comment
488
The comments are not visible in the normal build. Comment blocks wont be included in the output of
any build at all. Heres a comment block:
////
Note that includes in here will still be processed, but not make it into the output.
That is, missing includes here will still break the build!
////
30.3.12.Code Snippets
Explicitly defined in the document
Warning
Use this kind of code snippets as little as possible. They are well known to get out of sync
with reality after a while.
This is how to do it:
[source,cypher]
---start n=(2, 1) where (n.age < 30 and n.name = "Tobias") or not(n.name = "Tobias")
----
return n

start n=(2, 1) where (n.age < 30 and n.name = "Tobias") or not(n.name = "Tobias")
return n
If theres no suitable syntax highlighter, just omit the language: [source].

Currently the following syntax highlighters are enabled:
Bash
Cypher
Groovy
Java
JavaScript
Python
XML
For other highlighters we could add see http://alexgorbatchev.com/SyntaxHighlighter/manual/

brushes/.
Fetched from source code
Code can be automatically fetched from source files. You need to define:
component: the artifactId of the Maven coordinates,

source: path to the file inside the jar its deployed to,
classifier: sources or test-sources or any other classifier pointing to the artifact,
tag: tag name to search the file for,
the language of the code, if a corresponding syntax highlighter is available.
Note that the artifact has to be included as a Maven dependency of the Manual project so that the files
can be found.
Be aware of that the tag "abc" will match "abcd" as well. Its a simple on/off switch, meaning that
multiple occurrences will be assembled into a single code snippet in the output. This behavior can be
user to hide away assertions from code examples sourced from tests.
489
This is how to define a code snippet inclusion:
[snippet,java]
---component=neo4j-examples
source=org/neo4j/examples/JmxDocTest.java
classifier=test-sources
tag=getStartTime
----
This is how it renders:

private static Date getStartTimeFromManagementBean(
GraphDatabaseService graphDbService )
{
GraphDatabaseAPI graphDb = (GraphDatabaseAPI) graphDbService;
ObjectName objectName = JmxUtils.getObjectName( graphDb, "Kernel" );
Date date = JmxUtils.getAttribute( objectName, "KernelStartTime" );
return date;
}
Query Results
Theres a special filter for Cypher query results. This is how to tag a query result:
.Result
[queryresult]
---+----------------------------------+
| friend_of_friend.name | count(*) |
+----------------------------------+
| Ian
| 2
|
| Derrick
| 1
|
| Jill
| 1
|
+----------------------------------+
3 rows, 12 ms
----
This is how it renders:

Result
friend_of_friend.name
count(*)
Ian
Derrick
Jill
3 rows, 12 ms
30.3.13.A sample Java based documentation test

For Java, there are a couple of premade utilities that keep code and documentation together in
Javadocs and code snippets that generate Asciidoc for the rest of the toolchain.
To illustrate this, look at the following documentation that generates the Asciidoc file hello-worldtitle.asciidoc with a content of:
[[examples-hello-world-sample-chapter]]
Hello world Sample Chapter
==========================
490
This is a sample documentation test, demonstrating different ways of
bringing code and other artifacts into Asciidoc form. The title of the
generated document is determined from the method name, replacing "+_+" with
" ".
Below you see a number of different ways to generate text from source,
inserting it into the JavaDoc documentation (really being Asciidoc markup)
via the +@@+ snippet markers and programmatic adding with runtime data
in the Java code.
- The annotated graph as http://www.graphviz.org/[GraphViz]-generated visualization:
.Hello World Graph
["dot", "Hello-World-Graph-hello-world-Sample-Chapter.svg", "neoviz", ""]
---N0 [
label = "{Node\[0\]name = \'you\'\l}"
]
N1 [
label = "{Node\[1\]name = \'I\'\l}"
]
N1 -> N0 [
color = "2e3436"
fontcolor = "2e3436"
label = "know\n"
]
---- A sample Cypher query:
[source,cypher]
---START n = node(1)
RETURN n
---- A sample text output snippet:
[source]
---Hello graphy world!
---- a generated source link to the original GIThub source for this test:
https://github.com/neo4j/neo4j/blob/{neo4j-git-tag}/community/embedded-examples/src/test/java/org/neo4j/examples/DocumentationDocTest
- The full source for this example as a source snippet, highlighted as Java code:
[snippet,java]
---component=neo4j-examples
source=org/neo4j/examples/DocumentationDocTest.java
classifier=test-sources
tag=sampleDocumentation
---This is the end of this chapter.
this file is included in this documentation via

:leveloffset: 3
include::{importdir}/neo4j-examples-docs-jar/dev/examples/hello-world-sample-chapter.asciidoc[]
which renders the following chapter:

491
30.3.14.Hello world Sample Chapter
This is a sample documentation test, demonstrating different ways of bringing code and other artifacts
into Asciidoc form. The title of the generated document is determined from the method name,
replacing "_" with " ".
Below you see a number of different ways to generate text from source, inserting it into the JavaDoc
documentation (really being Asciidoc markup) via the @@ snippet markers and programmatic adding
with runtime data in the Java code.
The annotated graph as GraphViz <http://www.graphviz.org/>-generated visualization:
Figure30.1.Hello World Graph
know
A sample Cypher query:

START n = node(1)
RETURN n
A sample text output snippet:

Hello graphy world!
a generated source link to the original GIThub source for this test:
DocumentationDocTest.java <https://github.com/neo4j/neo4j/blob/2.0.0/community/embeddedexamples/src/test/java/org/neo4j/examples/DocumentationDocTest.java>
The full source for this example as a source snippet, highlighted as Java code:
// START SNIPPET: _sampleDocumentation
package org.neo4j.examples;
import org.junit.Test;
import org.neo4j.kernel.impl.annotations.Documented;
import org.neo4j.test.GraphDescription.Graph;
import static org.neo4j.visualization.asciidoc.AsciidocHelper.createGraphVizWithNodeId;
import static org.neo4j.visualization.asciidoc.AsciidocHelper.createOutputSnippet;
public class DocumentationDocTest extends ImpermanentGraphJavaDocTestBase
{
/**
* This is a sample documentation test, demonstrating different ways of
* bringing code and other artifacts into Asciidoc form. The title of the
* generated document is determined from the method name, replacing "+_+" with
* " ".
*
* Below you see a number of different ways to generate text from source,
492
* inserting it into the JavaDoc documentation (really being Asciidoc markup)
* via the +@@+ snippet markers and programmatic adding with runtime data
* in the Java code.
*
* - The annotated graph as http://www.graphviz.org/[GraphViz]-generated visualization:
*
* @@graph
*
* - A sample Cypher query:
*
* @@cypher
*
* - A sample text output snippet:
*
* @@output
*
* - a generated source link to the original GIThub source for this test:
*
* @@github
*
* - The full source for this example as a source snippet, highlighted as Java code:
*
* @@sampleDocumentation
*
* This is the end of this chapter.
*/
@Test
// signaling this to be a documentation test
@Documented
// the graph data setup as simple statements
@Graph( "I know you" )
// title is determined from the method name
public void hello_world_Sample_Chapter()
{
// initialize the graph with the annotation data
data.get();
gen.get().addTestSourceSnippets( this.getClass(), "sampleDocumentation" );
gen.get()
.addGithubTestSourceLink( "github", this.getClass(),
"community/embedded-examples" );
gen.get().addSnippet( "output",
createOutputSnippet( "Hello graphy world!" ) );
gen.get().addSnippet(
"graph",
createGraphVizWithNodeId( "Hello World Graph", graphdb(),
gen.get().getTitle() ) );
// A cypher snippet referring to the generated graph in the start clause
gen.get().addSnippet(
"cypher",
createCypherSnippet( "start n = node(" + data.get().get( "I" ).getId()
+ ") return n" ) );
}
}
// END SNIPPET: _sampleDocumentation
This is the end of this chapter.
30.3.15.Integrated Live Console

An interactive console can be added and will show up in the online HTML version. An optional title
can be added, which will be used for the text of the button.
This is how to do it, using Cypher to define the data, with an empty line to separate it from the query:
493
.Interactive Example
[console]
---CREATE (n {name: 'Neo4j'})
MATCH (n)
return n
----
And this is the result:
30.3.16.Toolchain
Useful links when configuring the docbook toolchain:
http://www.docbook.org/tdg/en/html/docbook.html
http://www.sagehill.net/docbookxsl/index.html
http://docbook.sourceforge.net/release/xsl/1.76.1/doc/html/index.html
http://docbook.sourceforge.net/release/xsl/1.76.1/doc/fo/index.html
494
30.4.Translating the Neo4j Manual

To translate the Neo4j Manual, theres a special project setup to use. See the French translation project
for an example: https://github.com/neo4j/manual-french
The project contains:
conf/ configuration for the project.

docs/ translated files for content provided by Neo4j modules.
po/ translation files and po4a configuration files.
src/ translated files for content provided by the original manual.
Makefile a makefile with project-specific configuration.
pom.xml Maven build configuration.
30.4.1.Prerequisites
Apache Maven
GNU Make
Python
Perl
Perl module: Unicode::GCString
To check if you have the Unicode::GCString module installed, you can issue the following command:
perl -MUnicode::GCString -e ''
If theres no error, the module has been successfully installed on your system.
To install the module, you can use cpanminus <http://search.cpan.org/dist/App-cpanminus/lib/App/
cpanminus.pm>. For a convenient way to install it, see http://cpanmin.us. With cpanminus installed,
execute this command:
cpanm Unicode::GCString
You will probably want to use a .po file editor as well, see Section30.4.5, Translation tools.
30.4.2.Build flow and file layout

The build is essentially a two-step process. The first step generates or copies translated documents,
while the second step is an ordinary AsciiDoc build using the output from the first step as sources.
Other than the src/ and docs/ diirectories of the project, the build generates files with the same layout
in two more places:
1. target/original/(src|docs)/ the contents of the original manual. Note thats it easier to look for
content here than to dig into the original manual itself.
2. target/(src|docs)/ the translated source to use for the AsciiDoc build.
The translated documents in target/(src|docs)/ are generated in three steps:
1. It starts out as a copy of the original manual.
2. Next, any static translated files fromt the src/ and docs/ directories of the project are copied.
3. Finally, the translation files in the po/ directory are used to generate translated documents.
495
Files produced by later steps will overwrite existing files from earlier steps.
30.4.3.Adding a chapter to a translation file

The translation is split over multiple translation files, one per part of the manual. Its all about
making the translation easier to manage and the tools to perform well. The basic rule of thumb is that
if some content is moved, it should likely still end up in the same translation file. In that case, the tools
will even detect this and the translation will be moved automatically.
To add a document to a translation file, do like this:
make add DOCUMENT="src/introduction/the-neo4j-graphdb.asciidoc" PART="introduction"
If the translation file does not already exist, it will be created. The document will be added to the
translation build configuration file as well. (The configuration is in the corresponding .conf file in the
po/ directory.)
If there exists a translated copy of the document at the location the DOCUMENT parameter points to, the
script will attempt to populate the translation file with translated paragraphs from that document.
Note that the structure of the document has to be a perfect match, or it will fail. However, the error
messages are helpful, so just fix and try again until it works! Translation file and configuration are
only changed when the first part succeeds.
Note
Only documents that need to be translated should be added. For example Cypher queries
and query results should not be translated. In general, documents residing in a directory
named includes should not be translated.
Also note that AsciiDoc include:: lines are normally not part of the translation at all, but handled
automatically. In case they need to be handled differently in a document, this has to be configured in
the corresponding .conf file. For example a normal document entry in such a file can look like this:
[type: asciidoc] target/original/src/operations/index.asciidoc fr:target/src/operations/index.asciidoc
To configure a single document not to handle include:: lines automatically, add the following at the
end of the line:
opt: "-o definitions=target/tools/main/resources/conf/translate-includes"
30.4.4.Workflow
First, use Maven to set up the environment and download the original manual and documentation
tools:
mvn clean package
To refresh the original manual and the tools, use the maven command again. For the sake of keeping
in sync with the original manual, a daily run of this command is recommended.
Once things are set up, use make during work.
make same as make preview.

make add add
a document to a translation file.

make preview refresh and build preview of the manual.
make refresh refresh translation files from the original and generated translated documents.
The preview of the translated manual is found in the target/html/ directory.

496
The actual work on translation is done by editing translation files. Suggested tools for that are found
below.
30.4.5.Translation tools
There are different editors for .po files containing the translations Below is a list of editors.
Gtranslator <http://projects.gnome.org/gtranslator/>
Lokalize <http://userbase.kde.org/Lokalize>
Virtaal <http://translate.sourceforge.net/wiki/virtaal/index>
Poedit <http://www.poedit.net/>
497
30.5.Contributing Code to Neo4j

30.5.1.Intro
The Neo4j community is a free software and open source community centered around software
and components for the Neo4j Graph Database. It is sponsored by Neo Technology <http://
neotechnology.com>, which provides infrastructure (different kinds of hosting, documentation, etc)
as well as people to manage it. The Neo4j community is an open community, in so far as it welcomes
any member that accepts the basic criterias of contribution and adheres to the communitys Code of
Conduct.
Contribution can be in many forms (documentation, discussions, bug reports). This document outlines
the rules of governance for a contributor of code.
30.5.2.Governance fundamentals
In a nutshell, you need to be aware of the following fundamentals if you wish to contribute code:
All software published by the Neo4j project must have been contributed under the Neo4j Code
Contributor License Agreement.
Neo4j is a free software and open source community. As a contributor, you are free to place your
work under any license that has been approved by either the Free Software Foundation <http://
www.fsf.org/> or the Open Source Initiative <http://opensource.org>. You still retain copyright, so
in addition to that license you can of course release your work under any other license (for example
a fully proprietary license), just not on the Neo4j infrastructure.
The Neo4j software is split into components. A Git repository holds either a single or multiple
components.
The source code should follow the Neo4j Code Style and fit in with the Neo4j infrastructure as
much as is reasonable for the specific component.
30.5.3.Contributor roles
Every individual that contributes code does so in the context of a role (a single individual can have
multiple roles). The role defines their responsibilities and privileges:
A patch submitter is a person who wishes to contribute a patch to an existing component. See
Workflow below.
A committer can contribute code directly to one or more components.
A component maintainer is in charge of a specific component. They can:
commit code in their components repository,
manage tickets for the repository,
grant push rights to the repository.
A Neo4j admin manages the Neo4j infrastructure. They:
define new components and assign component maintainership,
drive, mentor and coach Neo4j component development.
30.5.4.Contribution workflow
Code contributions to Neo4j are normally done via Github Pull Requests, following the workflow
shown below. Please check the pull request checklist before sending off a pull request as well.
1. Fork the appropriate Github repository.
498
2.
3.
4.
5.
6.
7.
Create a new branch for your specific feature or fix.

Write unit tests for your code.
Write code.
Write appropriate Javadocs and Manual entries.
Commit changes.
Send pull request.
30.5.5.Pull request checklist

1.
2.
3.
4.
5.
6.
Sign the CLA.

Ensure that you have not added any merge commits.
Squash all your changes into a single commit.
Rebase against the latest master.
Run all relevant tests.
Send the request!
30.5.6.Unit Tests
You have a much higher chance of getting your changes accepted if you supply us with small,
readable unit tests covering the code youve written. Also, make sure your code doesnt break
any existing tests. Note that there may be downstream components that need to be tested as well,
depending on what you change.
To run tests, use Maven rather than your IDE to ensure others can replicate your test run. The
command for running Neo4j tests in any given component is mvn clean validate.
30.5.7.Code Style
The Neo4j Code style is maintained on GitHub in styles for the different IDEs <https://github.com/
neo4j/neo4j.github.com/tree/master/code-style>.
30.5.8.Commit messages
Please take some care in providing good commit messages. Use your common sense. In particular:
Use english. This includes proper punctuation and correct spelling. Commit messages are supposed
to convey some information at a glance theyre not a chat room.
Remember that a commit is a changeset, which describes a cohesive set of changes across
potentially many files. Try to group every commit as a logical change. Explain what it changes. If
you have to redo work, you might want to clean up your commit log before doing a pull request.
If you fix a bug or an issue thats related to a ticket, then refer to the ticket in the message. For
example, Àdded this and then changed that. This fixes #14.' Just mentioning #xxx in the commit
will connect it to the GitHub issue with that number, see GitHub issues <https://github.com/
blog/831-issues-2-0-the-next-generation>. Any of these synonyms will also work:
fixes #xxx
fixed #xxx
fix #xxx
closes #xxx
close #xxx
closed #xxx.
499
Remember to convey intent. Dont be too brief but dont provide too much detail, either. Thats
what git diff is for.
30.5.9.Signing the CLA

One crucial aspect of contributing is the Contributor License Agreement. In short: make sure to sign
the CLA, or the Neo4j project wont be able to accept your contribution.
30.5.10.Dont merge, use rebase instead!

Because we would like each contribution to be contained in a single commit, merge commits are not
allowed inside a pull request. Merges are messy, and should only be done when necessary, eg. when
merging a branch into master to remember where the code came from.
If you want to update your development branch to incorporate the latest changes from master, use
git rebase. For details on how to use rebase, see Git manual on rebase: the Git Manual <http://gitscm.com/book/en/Git-Branching-Rebasing>.
30.5.11.Single commit
If you have multiple commits, you should squash them into a single one for the pull request, unless
there is some extraordinary reason not to. Keeping your changes in a single commit makes the commit
history easier to read, it also makes it easy to revert and move features around.
One way to do this is to, while standing on your local branch with your changes, create a new branch
and then interactively rebase your commits into a single one.
Interactive rebasing with Git.
On branch mychanges
git checkout -b mychanges-clean
Assuming you have 4 commits, rebase the last four commits interactively:
git rebase -i HEAD~4
In the dialog git gives you, keep your first commit, and squash all others into it.
Then reword the commit description to accurately depict what your commit does.
If applicable, include any issue numbers like so: 760
For more details, see the git manual: http://git-scm.com/book/en/Git-Tools-RewritingHistory#Changing-Multiple-Commit-Messages

If you are asked to modify parts of your code, work in your original branch (the one with multiple
commits), and follow the above process to create a fixed single commit.
500
30.6.Contributors
As an Open Source Project, the Neo4j User community extends its warmest thanks to all the
contributors who have signed the Section30.1, Contributor License Agreement to date and are
contributing to this collective effort.
name
GIThub ID
Johan Svensson
johan-neo <https://github.com/johan-neo>
Emil Eifrem
emileifrem <https://github.com/emileifrem>
Peter Neubauer
peterneubauer <https://github.com/peterneubauer>
Mattias Persson
tinwelint <https://github.com/tinwelint>
Tobias Lindaaker
thobe <https://github.com/thobe>
Anders Nawroth
nawroth <https://github.com/nawroth>
Andrs Taylor
systay <https://github.com/systay>
Jacob Hansson
jakewins <https://github.com/jakewins>
Jim Webber
jimwebber <https://github.com/jimwebber>
Josh Adell
jadell <https://github.com/jadell>
Andreas Kollegger
akollegger <https://github.com/akollegger>
Chris Gioran
digitalstain <https://github.com/digitalstain>
Thomas Baum
tbaum <https://github.com/tbaum>
Alistair Jones
apcj <https://github.com/apcj>
Michael Hunger
jexp <https://github.com/jexp>
Jesper Nilsson
jespernilsson <https://github.com/jespernilsson>
Tom Sulston
tomsulston <https://github.com/tomsulston>
David Montag
dmontag <https://github.com/dmontag>
Marlon Richert
marlonrichert <https://github.com/marlonrichert>
Hugo Josefson
hugojosefson <https://github.com/hugojosefson>
Vivek Prahlad
vivekprahlad <https://github.com/vivekprahlad>
Adriano Almeida
adrianoalmeida7 <https://github.com/
adrianoalmeida7>
Benjamin Gehrels
BGehrels <https://github.com/BGehrels>
Christopher Schmidt
FaKod <https://github.com/FaKod>
Pascal Rehfeldt
prehfeldt <https://github.com/prehfeldt>
Bjrn Sderqvist
cybear <https://github.com/cybear>
Abdul Azeez Shaik
abdulazeezsk <https://github.com/abdulazeezsk>
James Thornton
espeed <https://github.com/espeed>
Radhakrishna Kalyan
nrkkalyan <https://github.com/nrkkalyan>
Michel van den Berg
promontis <https://github.com/promontis>
Brandon McCauslin
bm3780 <https://github.com/bm3780>
501
name
GIThub ID
Hendy Irawan
ceefour <https://github.com/ceefour>
Luanne Misquitta
luanne <https://github.com/luanne>
Jim Radford
radford <https://github.com/radford>
Axel Morgner
amorgner <https://github.com/amorgner>
Taylor Buley
editor <https://github.com/editor>
Alex Smirnov
alexsmirnov <https://github.com/alexsmirnov>
Johannes Mockenhaupt
jotomo <https://github.com/jotomo>
Pablo Pareja Tobes
pablopareja <https://github.com/pablopareja>
Bjrn Granvik
bjorngranvik <https://github.com/bjorngranvik>
Julian Simpson
simpsonjulian <https://github.com/simpsonjulian>
Pablo Pareja Tobes
pablopareja <https://github.com/pablopareja>
Rickard berg
rickardoberg <https://github.com/rickardoberg>
Stefan Armbruster
sarmbruster <https://github.com/sarmbruster>
Stephan Hagemann
shageman <https://github.com/shageman>
Linan Wang
wangii <https://github.com/wangii>
Ian Robinson
iansrobinson <https://github.com/iansrobinson>
Marko Rodriguez
okram <https://github.com/okram>
Saikat Kanjilal
skanjila <https://github.com/skanjila>
Craig Taverner
craigtaverner <https://github.com/craigtaverner>
David Winslow
dwins <https://github.com/dwins>
Patrick Fitzgerald
paddydub <https://github.com/paddydub>
Stefan Berder
hrbonz <https://github.com/hrbonz>
Michael Kanner
SepiaGroup <https://github.com/SepiaGroup>
Lin Zhemin
miaoski <https://github.com/miaoski>
Christophe Willemsen
kwattro <https://github.com/kwattro>
Tony Liu
kooyeed <https://github.com/kooyeed>
Michael Klishin
michaelklishin <https://github.com/michaelklishin>
Wes Freeman
wfreeman <https://github.com/wfreeman>
Chris Leishman
cleishm <https://github.com/cleishm>
Brian Levine
blevine <https://github.com/blevine>
Ben Day
benday280412 <https://github.com/benday280412>
Davide Savazzi
svzdvd <https://github.com/svzdvd>
Nigel Small
nigelsmall <https://github.com/nigelsmall>
Lasse Westh-Nielsen
lassewesth <https://github.com/lassewesth>
Wujek Srujek
wujek-srujek <https://github.com/wujek-srujek>
502
name
GIThub ID
Alexander Yastrebov
AlexanderYastrebov <https://github.com/
AlexanderYastrebov>
Mike Bryant
mikesname <https://github.com/mikesname>
Klaus Grossmann
iKlaus <https://github.com/iKlaus>
Pablo Lalloni
plalloni <https://github.com/plalloni>
Stefan Plantikow
boggle <https://github.com/boggle>
Trenton Strong
trentonstrong <https://github.com/trentonstrong>
Maciej Mazur
mamciek <https://github.com/mamciek>
German Borbolla
germanborbolla <https://github.com/
germanborbolla>
Laurent Raufaste
lra <https://github.com/lra>
Thomas Hfele
Perfect-Pixel <https://github.com/Perfect-Pixel>
Sevki Hasirci
Sevki <https://github.com/Sevki>
Max De Marzi
maxdemarzi <https://github.com/maxdemarzi>
Pieter-Jan Van Aeken
PieterJanVanAeken <https://github.com/
PieterJanVanAeken>
Shane Gibbs
sgibbs-kellermed <https://github.com/sgibbskellermed>
Yin Wang
yinwang0 <https://github.com/yinwang0>
Volker Lanting
VolkerL <https://github.com/VolkerL>
Mark Needham
mneedham <https://github.com/mneedham>
Chris Vest
chrisvest <https://github.com/chrisvest>
Ben Butler-Cole
benbc <https://github.com/benbc>
Tatham Oddie
tathamoddie <https://github.com/tathamoddie>
503
PartVIII.Advanced Usage
This part contains information on advanced usage of Neo4j. Among the topics covered are embedding
Neo4j in your own software and writing plugins for the Neo4j Server.
Chapter31.Extending the Neo4j Server

The Neo4j Server can be extended by either plugins or unmanaged extensions.
505
Extending the Neo4j Server
31.1.Server Plugins
Quick info
The servers functionality can be extended by adding plugins.
Plugins are user-specified code which extend the capabilities of the database, nodes, or
relationships.
The neo4j server will then advertise the plugin functionality within representations as
clients interact via HTTP.
Plugins provide an easy way to extend the Neo4j REST API with new functionality, without the need
to invent your own API. Think of plugins as server-side scripts that can add functions for retrieving
and manipulating nodes, relationships, paths, properties or indices.
Tip
If you want to have full control over your API, and are willing to put in the effort, and
understand the risks, then Neo4j server also provides hooks for unmanaged extensions
based on JAX-RS.
The needed classes reside in the org.neo4j:server-api <http://search.maven.org/#search|gav|1|g
%3A%22org.neo4j%22%20AND%20a%3A%22server-api%22> jar file. See the linked page for
downloads and instructions on how to include it using dependency management. For Maven projects,
add the Server API dependencies in your pom.xml like this:
<dependency>
<artifactId>server-api</artifactId>
<version>${neo4j-version}</version>
</dependency>
Where ${neo4j-version} is the intended version.

To create a plugin, your code must inherit from the ServerPlugin <http://components.neo4j.org/serverapi/2.0.0/apidocs/org/neo4j/server/plugins/ServerPlugin.html> class. Your plugin should also:
ensure that it can produce an (Iterable of) Node, Relationship or Path, any Java primitive or String or
an instance of a org.neo4j.server.rest.repr.Representation
specify parameters,
specify a point of extension and of course
contain the application logic.
make sure that the discovery point type in the @PluginTarget and the @Source parameter are of the
same type.
An example of a plugin which augments the database (as opposed to nodes or relationships) follows:
Get all nodes or relationships plugin.
@Description( "An extension to the Neo4j Server for getting all nodes or relationships" )
public class GetAll extends ServerPlugin
{
@Name( "get_all_nodes" )
@Description( "Get all nodes from the Neo4j graph database" )
@PluginTarget( GraphDatabaseService.class )
public Iterable<Node> getAllNodes( @Source GraphDatabaseService graphDb )
{
506

return GlobalGraphOperations.at( graphDb ).getAllNodes();
}
@Description( "Get all relationships from the Neo4j graph database" )
@PluginTarget( GraphDatabaseService.class )
public Iterable<Relationship> getAllRelationships( @Source GraphDatabaseService graphDb )
{
return GlobalGraphOperations.at( graphDb ).getAllRelationships();
}
}
The full source code is found here: GetAll.java <https://github.com/neo4j/neo4j/blob/2.0.0/

community/server-examples/src/main/java/org/neo4j/examples/server/plugins/GetAll.java>
Find the shortest path between two nodes plugin.
public class ShortestPath extends ServerPlugin
{
@Description( "Find the shortest path between two nodes." )
@PluginTarget( Node.class )
public Iterable<Path> shortestPath(
@Source Node source,
@Description( "The node to find the shortest path to." )
@Parameter( name = "target" ) Node target,
@Description( "The relationship types to follow when searching for the shortest path(s). " +
"Order is insignificant, if omitted all types are followed." )
@Parameter( name = "types", optional = true ) String[] types,
@Description( "The maximum path length to search for, default value (if omitted) is 4." )
@Parameter( name = "depth", optional = true ) Integer depth )
{
Expander expander;
if ( types == null )
{
expander = Traversal.expanderForAllTypes();
}
else
{
expander = Traversal.emptyExpander();
for ( int i = 0; i < types.length; i++ )
{
expander = expander.add( DynamicRelationshipType.withName( types[i] ) );
}
}
PathFinder<Path> shortestPath = GraphAlgoFactory.shortestPath(
expander, depth == null ? 4 : depth.intValue() );
return shortestPath.findAllPaths( source, target );
}
}
The full source code is found here: ShortestPath.java <https://github.com/neo4j/neo4j/blob/2.0.0/

community/server-examples/src/main/java/org/neo4j/examples/server/plugins/ShortestPath.java>
To deploy the code, simply compile it into a .jar file and place it onto the server classpath (which by
convention is the plugins directory under the Neo4j server home directory).
Tip
Make sure the directories listings are retained in the jarfile by either building with default
Maven, or with jar -cvf myext.jar *, making sure to jar directories instead of specifying
single files.
The .jar file must include the file META-INF/services/org.neo4j.server.plugins.ServerPlugin with the
fully qualified name of the implementation class. This is an example with multiple entries, each on a
separate line:
507

org.neo4j.examples.server.plugins.GetAll
org.neo4j.examples.server.plugins.DepthTwo
org.neo4j.examples.server.plugins.ShortestPath
The code above makes an extension visible in the database representation (via the @PluginTarget
annotation) whenever it is served from the Neo4j Server. Simply changing the @PluginTarget
parameter to Node.class or Relationship.class allows us to target those parts of the data model
should we wish. The functionality extensions provided by the plugin are automatically advertised
in representations on the wire. For example, clients can discover the extension implemented by the
above plugin easily by examining the representations they receive as responses from the server, e.g.
by performing a GET on the default database URI:
curl -v http://localhost:7474/db/data/
The response to the GET request will contain (by default) a JSON container that itself contains a
container called "extensions" where the available plugins are listed. In the following case, we only
have the GetAll plugin registered with the server, so only its extension functionality is available.
Extension names will be automatically assigned, based on method names, if not specifically specified
using the @Name annotation.
{
"extensions-info" : "http://localhost:7474/db/data/ext",
"reference_node" : "http://localhost:7474/db/data/node/0",
"extensions" : {
"GetAll" : {
"get_all_nodes" : "http://localhost:7474/db/data/ext/GetAll/graphdb/get_all_nodes",
"get_all_relationships" : "http://localhost:7474/db/data/ext/GetAll/graphdb/getAllRelationships"
}
}
Performing a GET on one of the two extension URIs gives back the meta information about the service:
curl http://localhost:7474/db/data/ext/GetAll/graphdb/get_all_nodes
{
"extends" : "graphdb",
"description" : "Get all nodes from the Neo4j graph database",
"name" : "get_all_nodes",
"parameters" : [ ]
}
To use it, just POST to this URL, with parameters as specified in the description and encoded as
JSON data content. F.ex for calling the shortest path extension (URI gotten from a GET to http://
localhost:7474/db/data/node/123):
curl -X POST http://localhost:7474/db/data/ext/GetAll/node/123/shortestPath \
-H "Content-Type: application/json" \
-d '{"target":"http://localhost:7474/db/data/node/456&depth=5"}'
If everything is OK a response code 200 and a list of zero or more items will be returned. If nothing is
returned (null returned from extension) an empty result and response code 204 will be returned. If the
extension throws an exception response code 500 and a detailed error message is returned.
Extensions that do any kind of write operation will have to manage their own transactions, i.e.
transactions arent managed automatically.
Through this model, any plugin can naturally fit into the general hypermedia scheme that Neo4j
espouses meaning that clients can still take advantage of abstractions like Nodes, Relationships
508

and Paths with a straightforward upgrade path as servers are enriched with plugins (old clients dont
break).
509
31.2.Unmanaged Extensions
Quick info
Danger: Men at Work! The unmanaged extensions are a way of deploying arbitrary JAXRS code into the Neo4j server.
The unmanaged extensions are exactly that: unmanaged. If you drop poorly tested code into
the server, its highly likely youll degrade its performance, so be careful.
Some projects want extremely fine control over their server-side code. For this weve introduced an
unmanaged extension API.
Warning
This is a sharp tool, allowing users to deploy arbitrary JAX-RS <http://en.wikipedia.org/

wiki/JAX-RS> classes to the server and so you should be careful when thinking about
using this. In particular you should understand that its easy to consume lots of heap space
on the server and hinder performance if youre not careful.
Still, if you understand the disclaimer, then you load your JAX-RS classes into the Neo4j server
simply by adding a @Context annotation to your code, compiling against the JAX-RS jar and any
Neo4j jars youre making use of. Then add your classes to the runtime classpath (just drop it in the lib
directory of the Neo4j server). In return you get access to the hosted environment of the Neo4j server
like logging through the org.neo4j.server.logging.Logger.
In your code, you get access to the underlying GraphDatabaseService through the @Context annotation
like so:
public MyCoolService( @Context GraphDatabaseService database )
{
// Have fun here, but be safe!
}
Remember, the unmanaged API is a very sharp tool. Its all to easy to compromise the server by
deploying code this way, so think first and see if you cant use the managed extensions in preference.
However, a number of context parameters can be automatically provided for you, like the reference to
the database.
In order to specify the mount point of your extension, a full class looks like this:
Unmanaged extension example.
@Path( "/helloworld" )
public class HelloWorldResource
{
private final GraphDatabaseService database;
public HelloWorldResource( @Context GraphDatabaseService database )
{
this.database = database;
}
@GET
@Produces( MediaType.TEXT_PLAIN )
@Path( "/{nodeId}" )
public Response hello( @PathParam( "nodeId" ) long nodeId )
{
510

// Do stuff with the database
return Response.status( Status.OK ).entity(
("Hello World, nodeId=" + nodeId).getBytes( Charset.forName("UTF-8") ) ).build();
}
}
The full source code is found here: HelloWorldResource.java <https://github.com/neo4j/neo4j/

blob/2.0.0/community/server-examples/src/main/java/org/neo4j/examples/server/unmanaged/
HelloWorldResource.java>
Build this code, and place the resulting jar file (and any custom dependencies) into the
$NEO4J_SERVER_HOME/plugins directory, and include this class in the neo4j-server.properties file, like so:
Tip
Make sure the directories listings are retained in the jarfile by either building with default
Maven, or with jar -cvf myext.jar *, making sure to jar directories instead of specifying
single files.
Tip
You will need to include a dependency to JAX-RS API on your classpath when you
compile. In Maven this would be achieved by adding the following to the pom file:
<dependency>
<groupId>javax.ws.rs</groupId>
<artifactId>javax.ws.rs-api</artifactId>
<version>2.0</version>
<scope>provided</scope>
</dependency>
Comma separated list of JAXRS packages containing JAXRS Resource, one package name for each mountpoint.
org.neo4j.server.thirdparty_jaxrs_classes=org.neo4j.examples.server.unmanaged=/examples/unmanaged
Which binds the hello method to respond to GET requests at the URI: http://{neo4j_server}:
{neo4j_port}/examples/unmanaged/helloworld/{nodeId}
curl http://localhost:7474/examples/unmanaged/helloworld/123
which results in
Hello World, nodeId=123
511
31.3.Installing Plugins and Extensions in Neo4j Desktop

Neo4j Desktop can also be extended with server plugins and extensions. Neo4j Desktop will add
all jars in %ProgramFiles%\Neo4j Community\plugins to the classpath, but please note that nested
directories for plugins are currently not supported.
Otherwise server plugins and extensions are subject to the same rules as usual. Please note when
configuring server extensions that neo4j-server.properties for Neo4j Desktop lives in %APPDATA%
\Neo4j Community.
512
Chapter32.Using Neo4j embedded in Java applications

Its easy to use Neo4j embedded in Java applications. In this chapter you will find everything
needed from setting up the environment to doing something useful with your data.
513
Using Neo4j embedded in Java applications
32.1.Include Neo4j in your project
After selecting the appropriate edition for your platform, embed Neo4j in your Java application by
including the Neo4j library jars in your build. The following sections will show how to do this by
either altering the build path directly or by using dependency management.
32.1.1.Add Neo4j to the build path

Get the Neo4j libraries from one of these sources:
Extract a Neo4j download <http://neo4j.org/download/> zip/tarball, and use the jar files found in
the lib/ directory.
Use the jar files available from Maven Central Repository <http://search.maven.org/#search|ga|1|g
%3A%22org.neo4j%22>
Add the jar files to your project:
JDK tools
Append to -classpath
Eclipse
Right-click on the project and then go Build Path Configure Build Path. In the dialog,
choose Add External JARs, browse to the Neo4j lib/ directory and select all of the jar files.
Another option is to use User Libraries <http://help.eclipse.org/indigo/index.jsp?topic=/
org.eclipse.jdt.doc.user/reference/preferences/java/buildpath/ref-preferences-userlibraries.htm>.
IntelliJ IDEA
See Libraries, Global Libraries, and the Configure Library dialog <http://www.jetbrains.com/idea/
webhelp/configuring-project-and-global-libraries.html>
NetBeans
Right-click on the Libraries node of the project, choose Add JAR/Folder, browse to the Neo4j
lib/ directory and select all of the jar files.
You can also handle libraries from the project node, see Managing a Projects Classpath <http://
netbeans.org/kb/docs/java/project-setup.html#projects-classpath>.
32.1.2.Editions
The following table outlines the available editions and their names for use with dependency
management tools.
Tip
Follow the links in the table for details on dependency configuration with Apache Maven,
Apache Buildr, Apache Ivy, Groovy Grape, Grails, Scala SBT!
Neo4j editions
Edition
Community
Dependency
Description
License
org.neo4j:neo4j <http://
search.maven.org/
#search|gav|1|g
%3A%22org.neo4j
%22%20AND%20a
%3A%22neo4j%22>
a high performance,
fully ACID
transactional graph
database
GPLv3
514

Edition
Dependency
Description
License
Enterprise
org.neo4j:neo4jenterprise <http://
search.maven.org/
#search|gav|1|g
%3A%22org.neo4j
%22%20AND%20a
%3A%22neo4jenterprise%22>
adding advanced
monitoring, online
backup and High
Availability clustering
AGPLv3
Note
The listed dependencies do not contain the implementation, but pulls it in transitively.
For more information regarding licensing, see the Licensing Guide <http://www.neo4j.org/learn/
licensing>.
32.1.3.Add Neo4j as a dependency

You can either go with the top-level artifact from the table above or include the individual
components directly. The examples included here use the top-level artifact approach.
Maven
Maven dependency.
<project>
...
<dependencies>
<dependency>
<artifactId>neo4j</artifactId>
<version>2.0.0</version>
</dependency>
...
</dependencies>
...
</project>
Where the artifactId is found in the editions table.

Eclipse and Maven
For development in Eclipse <http://www.eclipse.org>, it is recommended to install the m2e plugin
<http://www.eclipse.org/m2e/> and let Maven manage the project build classpath instead, see above.
This also adds the possibility to build your project both via the command line with Maven and have a
working Eclipse setup for development.
Ivy
Make sure to resolve dependencies from Maven Central, for example using this configuration in your
ivysettings.xml file:
<ivysettings>
<settings defaultResolver="main"/>
<resolvers>
<chain name="main">
<filesystem name="local">
<artifact pattern="${ivy.settings.dir}/repository/[artifact]-[revision].[ext]" />
515

</filesystem>
<ibiblio name="maven_central" root="http://repo1.maven.org/maven2/" m2compatible="true"/>
</chain>
</resolvers>
</ivysettings>
With that in place you can add Neo4j to the mix by having something along these lines to your ivy.xml
file:
..
<dependencies>
..
<dependency org="org.neo4j" name="neo4j" rev="2.0.0"/>
..
</dependencies>
..
Where the name is found in the editions table above

Gradle
The example below shows an example gradle build script for including the Neo4j libraries.
def neo4jVersion = "2.0.0"
apply plugin: 'java'
repositories {
mavenCentral()
}
dependencies {
compile "org.neo4j:neo4j:${neo4jVersion}"
}
Where the coordinates (org.neo4j:neo4j in the example) are found in the editions table above.
32.1.4.Starting and stopping

To create a new database or pen an existing one you instantiate an EmbeddedGraphDatabase <http://
components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/kernel/EmbeddedGraphDatabase.html>.
graphDb = new GraphDatabaseFactory().newEmbeddedDatabase( DB_PATH );
registerShutdownHook( graphDb );
Note
The EmbeddedGraphDatabase instance can be shared among multiple threads. Note however
that you cant create multiple instances pointing to the same database.
To stop the database, call the shutdown() method:
graphDb.shutdown();
To make sure Neo4j is shut down properly you can add a shutdown hook:
private static void registerShutdownHook( final GraphDatabaseService graphDb )
{
// Registers a shutdown hook for the Neo4j instance so that it
// shuts down nicely when the VM exits (even if you "Ctrl-C" the
// running application).
Runtime.getRuntime().addShutdownHook( new Thread()
{
@Override
public void run()
{
graphDb.shutdown();
}
516

} );
}
Starting an embedded database with configuration settings

To start Neo4j with configuration settings, a Neo4j properties file can be loaded like this:
GraphDatabaseService graphDb = new GraphDatabaseFactory()
.newEmbeddedDatabaseBuilder( storeDir )
.loadPropertiesFromFile( pathToConfig + "neo4j.properties" )
.newGraphDatabase();
Or you could of course create you own Map<String, String> programmatically and use that instead,
like so:
Map<String, String> config = new HashMap<String, String>();
config.put( "neostore.nodestore.db.mapped_memory", "10M" );
config.put( "string_block_size", "60" );
config.put( "array_block_size", "300" );
GraphDatabaseService graphDb = new GraphDatabaseFactory()
.newEmbeddedDatabaseBuilder( storeDir )
.setConfig( config )
For configuration settings, see Chapter22, Configuration & Performance.

Starting an embedded read-only instance
If you want a read-only view of the database, create an instance this way:
Map<String, String> config = new HashMap<>();
config.put( "read_only", "true" );
graphDb = new GraphDatabaseFactory().newEmbeddedDatabaseBuilder(
"target/read-only-db/location" )
Obviously the database has to already exist in this case.
Note
Concurrent access to the same database files by multiple (read-only or write) instances is
not supported.
517
32.2.Hello World
Learn how to create and access nodes and relationships. For information on project setup, see
Section32.1, Include Neo4j in your project.
Remember, from Section2.1, What is a Graph Database?, that a Neo4j graph consist of:
Nodes that are connected by
Relationships, with
Properties on both nodes and relationships.
All relationships have a type. For example, if the graph represents a social network, a relationship type
could be KNOWS. If a relationship of the type KNOWS connects two nodes, that probably represents two
people that know each other. A lot of the semantics (that is the meaning) of a graph is encoded in the
relationship types of the application. And although relationships are directed they are equally well
traversed regardless of which direction they are traversed.
Tip
The source code of this example is found here: EmbeddedNeo4j.java <https://github.com/

neo4j/neo4j/blob/2.0.0/community/embedded-examples/src/main/java/org/neo4j/examples/
EmbeddedNeo4j.java>
32.2.1.Prepare the database

Relationship types can be created by using an enum. In this example we only need a single relationship
type. This is how to define it:
private static enum RelTypes implements RelationshipType
{
KNOWS
}
We also prepare some variables to use:

GraphDatabaseService graphDb;
Node firstNode;
Node secondNode;
Relationship relationship;
The next step is to start the database server. Note that if the directory given for the database doesnt
already exist, it will be created.
registerShutdownHook( graphDb );
Note that starting a database server is an expensive operation, so dont start up a new instance
every time you need to interact with the database! The instance can be shared by multiple threads.
Transactions are thread confined.
As seen, we register a shutdown hook that will make sure the database shuts down when the JVM
exits. Now its time to interact with the database.
32.2.2.Wrap operations in a transaction

All operations have to be performed in a transaction. This is a conscious design decision, since we
believe transaction demarcation to be an important part of working with a real enterprise database.
Now, transaction handling in Neo4j is very easy:
try ( Transaction tx = graphDb.beginTx() )
518

{
// Database operations go here
tx.success();
}
For more information on transactions, see Chapter16, Transaction Management and Java API for
Transaction <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/Transaction.html>.
Note
For brevity, we do not spell out wrapping of operations in a transaction throughout the
manual.
32.2.3.Create a small graph

Now, lets create a few nodes. The API is very intuitive. Feel free to have a look at the JavaDocs
at http://components.neo4j.org/neo4j/2.0.0/apidocs/. Theyre included in the distribution, as well.
Heres how to create a small graph consisting of two nodes, connected with one relationship and some
properties:
firstNode = graphDb.createNode();
firstNode.setProperty( "message", "Hello, " );
secondNode = graphDb.createNode();
secondNode.setProperty( "message", "World!" );
relationship = firstNode.createRelationshipTo( secondNode, RelTypes.KNOWS );
relationship.setProperty( "message", "brave Neo4j " );
We now have a graph that looks like this:

Figure32.1.Hello World Graph
m essage = 'Hello, '
KNOWS
m essage = 'brave Neo4j '
m essage = 'World! '
32.2.4.Print the result

After weve created our graph, lets read from it and print the result.
System.out.print( firstNode.getProperty( "message" ) );
System.out.print( relationship.getProperty( "message" ) );
System.out.print( secondNode.getProperty( "message" ) );
Which will output:

Hello, brave Neo4j World!
32.2.5.Remove the data

In this case well remove the data before committing:
// let's remove the data
firstNode.getSingleRelationship( RelTypes.KNOWS, Direction.OUTGOING ).delete();
519

firstNode.delete();
secondNode.delete();
Note that deleting a node which still has relationships when the transaction commits will fail. This is
to make sure relationships always have a start node and an end node.
32.2.6.Shut down the database server

Finally, shut down the database server when the application finishes:
graphDb.shutdown();
520
32.3.User database with legacy index

Note
Please read Section32.5, Managing resources when using long running transactions on
how to properly close ResourceIterators returned from index lookups.
You have a user database, and want to retrieve users by name using the legacy indexing system.
Tip
The source code used in this example is found here: EmbeddedNeo4jWithIndexing.java

<https://github.com/neo4j/neo4j/blob/2.0.0/community/embedded-examples/src/main/java/
org/neo4j/examples/EmbeddedNeo4jWithIndexing.java>
We have created two helper methods to handle user names and adding users to the database:
private static String idToUserName( final int id )
{
return "user" + id + "@neo4j.org";
}
private static Node createAndIndexUser( final String username )
{
Node node = graphDb.createNode();
node.setProperty( USERNAME_KEY, username );
nodeIndex.add( node, USERNAME_KEY, username );
return node;
}
The next step is to start the database server:

registerShutdownHook();
Its time to add the users:

{
nodeIndex = graphDb.index().forNodes( "nodes" );
// Create some users and index their names with the IndexService
for ( int id = 0; id < 100; id++ )
{
createAndIndexUser( idToUserName( id ) );
}
And heres how to find a user by Id:

int idToFind = 45;
String userName = idToUserName( idToFind );
Node foundUser = nodeIndex.get( USERNAME_KEY, userName ).getSingle();
System.out.println( "The username of user " + idToFind + " is "
+ foundUser.getProperty( USERNAME_KEY ) );
521
32.4.User database with indexes

You have a user database, and want to retrieve users by name using indexes.
Tip
The source code used in this example is found here:

EmbeddedNeo4jWithNewIndexing.java <https://github.com/neo4j/neo4j/
blob/2.0.0/community/embedded-examples/src/main/java/org/neo4j/examples/
EmbeddedNeo4jWithNewIndexing.java>
To begin with, we start the database server:
GraphDatabaseService graphDb = new GraphDatabaseFactory().newEmbeddedDatabase( DB_PATH );
Then we have to configure the database to index users by name. This only needs to be done once.
IndexDefinition indexDefinition;
{
Schema schema = graphDb.schema();
indexDefinition = schema.indexFor( DynamicLabel.label( "User" ) )
.on( "username" )
.create();
tx.success();
}
Indexes are populated asynchronously when they are first created. It is possible to use the core API to
wait for index population to complete:
{
Schema schema = graphDb.schema();
schema.awaitIndexOnline( indexDefinition, 10, TimeUnit.SECONDS );
}
Its time to add the users:

{
Label label = DynamicLabel.label( "User" );
// Create some users
for ( int id = 0; id < 100; id++ )
{
Node userNode = graphDb.createNode( label );
userNode.setProperty( "username", "user" + id + "@neo4j.org" );
}
System.out.println( "Users created" );
tx.success();
}
And heres how to find a user by id:
Note
Please read Section32.5, Managing resources when using long running transactions on
how to properly close ResourceIterators returned from index lookups.
int idToFind = 45;
String nameToFind = "user" + idToFind + "@neo4j.org";
{
522

try ( ResourceIterator<Node> users =
graphDb.findNodesByLabelAndProperty( label, "username", nameToFind ).iterator() )
{
ArrayList<Node> userNodes = new ArrayList<>();
while ( users.hasNext() )
{
userNodes.add( users.next() );
}
for ( Node node : userNodes )
{
System.out.println( "The username of user " + idToFind + " is " + node.getProperty( "username" ) );
}
}
}
When updating the name of a user, the index is updated as well:

{
int idToFind = 45;
for ( Node node : graphDb.findNodesByLabelAndProperty( label, "username", nameToFind ) )
{
node.setProperty( "username", "user" + ( idToFind + 1 ) + "@neo4j.org" );
}
tx.success();
}
When deleting a user, it is automatically removed from the index:

{
int idToFind = 46;
for ( Node node : graphDb.findNodesByLabelAndProperty( label, "username", nameToFind ) )
{
node.delete();
}
tx.success();
}
In case we change our data model, we can drop the index as well:
{
for ( IndexDefinition indexDefinition : graphDb.schema()
.getIndexes( label ) )
{
// There is only one index
indexDefinition.drop();
}
tx.success();
}
523
32.5.Managing resources when using long running

transactions
It is necessary to always open a transaction when accessing the database. Inside a long running
transaction it is good practice to ensure that any ResourceIterator <http://components.neo4j.org/
neo4j/2.0.0/apidocs/org/neo4j/graphdb/ResourceIterator.html>s obtained inside the transaction are
closed as early as possible. This is either achieved by just exhausting the iterator or by explicitly
calling its close method.
What follows is an example of how to work with a ResourceIterator. As we dont exhaust the iterator,
we will close it explicitly using the close() method.
int idToFind = 45;
try ( Transaction tx = graphDb.beginTx();
ResourceIterator<Node> users = graphDb
.findNodesByLabelAndProperty( label, "username", nameToFind )
.iterator() )
{
Node firstUserNode;
if ( users.hasNext() )
{
firstUserNode = users.next();
}
users.close();
}
524
32.6.Basic unit testing

The basic pattern of unit testing with Neo4j is illustrated by the following example.
To access the Neo4j testing facilities you should have the neo4j-kernel tests.jar on the classpath
during tests. You can download it from Maven Central: org.neo4j:neo4j-kernel <http://
search.maven.org/#search|ga|1|g%3A%22org.neo4j%22%20AND%20a%3A%22neo4j-kernel%22>.
Using Maven as a dependency manager you would typically add this dependency together with JUnit
and Hamcrest like so:
Maven dependency.
<project>
...
<dependencies>
<dependency>
<artifactId>neo4j-kernel</artifactId>
<version>2.0.0</version>
<type>test-jar</type>
<scope>test</scope>
</dependency>
<dependency>
<groupId>junit</groupId>
<artifactId>junit-dep</artifactId>
<scope>test</scope>
</dependency>
<dependency>
<groupId>org.hamcrest</groupId>
<artifactId>hamcrest-all</artifactId>
<scope>test</scope>
</dependency>
...
</dependencies>
...
</project>
Observe that the <type>test-jar</type> is crucial. Without it you would get the common neo4j-kernel
jar, not the one containing the testing facilities.
With that in place, were ready to code our tests.
Tip
For the full source code of this example see: Neo4jBasicDocTest.java <https://github.com/
neo4j/neo4j/blob/2.0.0/community/embedded-examples/src/test/java/org/neo4j/examples/
Neo4jBasicDocTest.java>
Before each test, create a fresh database:
@Before
public void prepareTestDatabase()
{
graphDb = new TestGraphDatabaseFactory().newImpermanentDatabase();
}
After the test has executed, the database should be shut down:
@After
public void destroyTestDatabase()
{
525

graphDb.shutdown();
}
During a test, create nodes and check to see that they are there, while enclosing write operations in a
transaction.
Node n = null;
{
n = graphDb.createNode();
n.setProperty( "name", "Nancy" );
tx.success();
}
// The node should have a valid id
assertThat( n.getId(), is( greaterThan( -1L ) ) );
// Retrieve a node by using the id of the created node. The id's and
// property should match.
{
Node foundNode = graphDb.getNodeById( n.getId() );
assertThat( foundNode.getId(), is( n.getId() ) );
assertThat( (String) foundNode.getProperty( "name" ), is( "Nancy" ) );
}
If you want to set configuration parameters at database creation, its done like this:
Map<String, String> config = new HashMap<String, String>();
config.put( "string_block_size", "60" );
config.put( "array_block_size", "300" );
GraphDatabaseService db = new TestGraphDatabaseFactory()
.newImpermanentDatabaseBuilder()
526
32.7.Traversal
For reading about traversals, see Chapter33, The Traversal Framework.
For more examples of traversals, see Chapter5, Data Modeling Examples.
32.7.1.The Matrix
This is the first graph we want to traverse into:
Figure32.2.Matrix node space view
Tip
The source code of this example is found here: NewMatrix.java <https://github.com/neo4j/

neo4j/blob/2.0.0/community/embedded-examples/src/main/java/org/neo4j/examples/
NewMatrix.java>
Friends and friends of friends.
private static Traverser getFriends(
final Node person )
{
TraversalDescription td = Traversal.description()
.breadthFirst()
.relationships( RelTypes.KNOWS, Direction.OUTGOING )
.evaluator( Evaluators.excludeStartPosition() );
return td.traverse( person );
}
Lets perform the actual traversal and print the results:

int numberOfFriends = 0;
String output = neoNode.getProperty( "name" ) + "'s friends:\n";
Traverser friendsTraverser = getFriends( neoNode );
for ( Path friendPath : friendsTraverser )
{
output += "At depth " + friendPath.length() + " => "
+ friendPath.endNode()
.getProperty( "name" ) + "\n";
numberOfFriends++;
}
527

output += "Number of friends found: " + numberOfFriends + "\n";
Which will give us the following output:

Thomas Anderson's friends:
At depth 1 => Trinity
At depth 1 => Morpheus
At depth 2 => Cypher
At depth 3 => Agent Smith
Number of friends found: 4
Who coded the Matrix?

private static Traverser findHackers( final Node startNode )
{
.breadthFirst()
.relationships( RelTypes.CODED_BY, Direction.OUTGOING )
.relationships( RelTypes.KNOWS, Direction.OUTGOING )
.evaluator(
Evaluators.includeWhereLastRelationshipTypeIs( RelTypes.CODED_BY ) );
return td.traverse( startNode );
}
Print out the result:

String output = "Hackers:\n";
int numberOfHackers = 0;
Traverser traverser = findHackers( getNeoNode() );
for ( Path hackerPath : traverser )
{
output += "At depth " + hackerPath.length() + " => "
+ hackerPath.endNode()
.getProperty( "name" ) + "\n";
numberOfHackers++;
}
output += "Number of hackers found: " + numberOfHackers + "\n";
Now we know who coded the Matrix:

Hackers:
At depth 4 => The Architect
Number of hackers found: 1
Walking an ordered path

This example shows how to use a path context holding a representation of a path.
Tip
The source code of this example is found here: OrderedPath.java <https://github.com/

neo4j/neo4j/blob/2.0.0/community/embedded-examples/src/main/java/org/neo4j/examples/
orderedpath/OrderedPath.java>
Create a toy graph.
Node
Node
Node
Node
A
B
C
D
=
=
=
=
db.createNode();
db.createNode();
db.createNode();
db.createNode();
A.createRelationshipTo(
C.createRelationshipTo(
A.createRelationshipTo(
B.createRelationshipTo(
C,
D,
B,
C,
REL2
REL3
REL1
REL2
);
);
);
);
528
REL1
REL2
REL2
REL3
Now, the order of relationships (REL1 REL2 REL3) is stored in an ArrayList. Upon traversal,
the Evaluator can check against it to ensure that only paths are included and returned that have the
predefined order of relationships:
Define how to walk the path.
final ArrayList<RelationshipType> orderedPathContext = new ArrayList<RelationshipType>();
orderedPathContext.add( REL1 );
orderedPathContext.add( withName( "REL2" ) );
orderedPathContext.add( withName( "REL3" ) );
.evaluator( new Evaluator()
{
@Override
public Evaluation evaluate( final Path path )
{
if ( path.length() == 0 )
{
return Evaluation.EXCLUDE_AND_CONTINUE;
}
RelationshipType expectedType = orderedPathContext.get( path.length() - 1 );
boolean isExpectedType = path.lastRelationship()
.isType( expectedType );
boolean included = path.length() == orderedPathContext.size() && isExpectedType;
boolean continued = path.length() < orderedPathContext.size() && isExpectedType;
return Evaluation.of( included, continued );
}
} )
.uniqueness( Uniqueness.NODE_PATH );
Note that we set the uniqueness to Uniqueness.NODE_PATH <http://components.neo4j.org/neo4j/2.0.0/

apidocs/org/neo4j/kernel/Uniqueness.htmlNODE_PATH> as we want to be able to revisit the same node
dureing the traversal, but not the same path.
Perform the traversal and print the result.
529

Traverser traverser = td.traverse( A );
PathPrinter pathPrinter = new PathPrinter( "name" );
for ( Path path : traverser )
{
output += Traversal.pathToString( path, pathPrinter );
}
Which will output:

(A)--[REL1]-->(B)--[REL2]-->(C)--[REL3]-->(D)
In this case we use a custom class to format the path output. This is how its done:
static class PathPrinter implements Traversal.PathDescriptor<Path>
{
private final String nodePropertyKey;
public PathPrinter( String nodePropertyKey )
{
this.nodePropertyKey = nodePropertyKey;
}
@Override
public String nodeRepresentation( Path path, Node node )
{
return "(" + node.getProperty( nodePropertyKey, "" ) + ")";
}
@Override
public String relationshipRepresentation( Path path, Node from, Relationship relationship )
{
String prefix = "--", suffix = "--";
if ( from.equals( relationship.getEndNode() ) )
{
prefix = "<--";
}
else
{
suffix = "-->";
}
return prefix + "[" + relationship.getType().name() + "]" + suffix;
}
}
For options regarding output of a Path, see the Traversal <http://components.neo4j.org/neo4j/2.0.0/

apidocs/org/neo4j/kernel/Traversal.html> class.
32.7.2.Uniqueness of Paths in traversals
This example is demonstrating the use of node uniqueness. Below an imaginary domain graph with
Principals that own pets that are descendant to other pets.
Figure32.3.Descendants Example Graph
Node[ 4] nam e = 'Principal1'
owns
owns
Node[ 0] nam e = 'Pet 1'
descendant
descendant
530
Node[ 5] nam e = 'Principal2'
descendant
owns

In order to return all descendants of Pet0 which have the relation owns to Principal1 (Pet1 and Pet3),
the Uniqueness of the traversal needs to be set to NODE_PATH rather than the default NODE_GLOBAL so that
nodes can be traversed more that once, and paths that have different nodes but can have some nodes in
common (like the start and end node) can be returned.
final Node target = data.get().get( "Principal1" );
.uniqueness( Uniqueness.NODE_PATH )
.evaluator( new Evaluator()
{
@Override
public Evaluation evaluate( Path path )
{
boolean endNodeIsTarget = path.endNode().equals( target );
return Evaluation.of( endNodeIsTarget, !endNodeIsTarget );
}
} );
Traverser results = td.traverse( start );
This will return the following paths:

(2)--[descendant,0]-->(0)<--[owns,3]--(4)
(2)--[descendant,2]-->(3)<--[owns,5]--(4)
In the default path.toString() implementation, (1)--[knows,2]-->(4) denotes a node with ID=1 having
a relationship with ID 2 or type knows to a node with ID-4.
Lets create a new TraversalDescription from the old one, having NODE_GLOBAL uniqueness to see the
difference.
Tip
The TraversalDescription object is immutable, so we have to use the new instance returned
with the new uniqueness setting.
TraversalDescription nodeGlobalTd = td.uniqueness( Uniqueness.NODE_GLOBAL );
results = nodeGlobalTd.traverse( start );
Now only one path is returned:

(2)--[descendant,0]-->(0)<--[owns,3]--(4)
32.7.3.Social network
Note
The following example uses the new enhanced traversal API.

Social networks (know as social graphs out on the web) are natural to model with a graph. This
example shows a very simple social model that connects friends and keeps track of status updates.
Tip
The source code of the example is found here: socnet <https://github.com/neo4j/neo4j/

tree/2.0.0/community/embedded-examples/src/main/java/org/neo4j/examples/socnet>
531

Simple social model
Figure32.4.Social network data model
The data model for a social network is pretty simple: Persons with names and StatusUpdates with
timestamped text. These entities are then connected by specific relationships.
Person
friend: relates two distinct Person instances (no self-reference)
status: connects to the most recent StatusUpdate
StatusUpdate
next: points to the next StatusUpdate in the chain, which was posted before the current one
Status graph instance
The StatusUpdate list for a Person is a linked list. The head of the list (the most recent status) is found
by following status. Each subsequent StatusUpdate is connected by next.
Heres an example where Andreas Kollegger micro-blogged his way to work in the morning:
532
Andreas Kollegger
st at us
st art ed designing t his graph m odel

9:30 am
next
rode m y awesom e Skeppshult t o work

8:45 am
next
is get t ing used t o m uesli for breakfast

8:00 am
To read the status updates, we can create a traversal, like so:

TraversalDescription traversal = Traversal.description().
depthFirst().
relationships( NEXT );
This gives us a traverser that will start at one StatusUpdate, and will follow the chain of updates until
they run out. Traversers are lazy loading, so its performant even when dealing with thousands of
statuses they are not loaded until we actually consume them.
Activity stream
Once we have friends, and they have status messages, we might want to read our friends status'
messages, in reverse time order latest first. To do this, we go through these steps:
1.
2.
3.
4.
5.
Gather all friends status update iterators in a list latest date first.
Sort the list.
Return the first item in the list.
If the first iterator is exhausted, remove it from the list. Otherwise, get the next item in that iterator.
Go to step 2 until there are no iterators left in the list.
Animated, the sequence looks like this <http://www.slideshare.net/systay/pattern-activity-stream>.

The code looks like:
PositionedIterator<StatusUpdate> first = statuses.get(0);
StatusUpdate returnVal = first.current();
if ( !first.hasNext() )
{
statuses.remove( 0 );
}
else
533

{
first.next();
sort();
}
return returnVal;
534
32.8.Domain entities
This page demonstrates one way to handle domain entities when using Neo4j. The principle at use is
to wrap the entities around a node (the same approach can be used with relationships as well).
Tip
The source code of the examples is found here: Person.java <https://github.com/neo4j/

neo4j/blob/2.0.0/community/embedded-examples/src/main/java/org/neo4j/examples/
socnet/Person.java>
First off, store the node and make it accessible inside the package:
private final Node underlyingNode;
Person( Node personNode )
{
this.underlyingNode = personNode;
}
protected Node getUnderlyingNode()
{
return underlyingNode;
}
Delegate attributes to the node:

public String getName()
{
return (String)underlyingNode.getProperty( NAME );
}
Make sure to override these methods:

@Override
{
return underlyingNode.hashCode();
}
@Override
public boolean equals( Object o )
{
return o instanceof Person &&
underlyingNode.equals( ( (Person)o ).getUnderlyingNode() );
}
@Override
public String toString()
{
return "Person[" + getName() + "]";
}
535
32.9.Graph Algorithm examples

Tip
The source code used in the example is found here: PathFindingDocTest.java <https://
github.com/neo4j/neo4j/blob/2.0.0/community/embedded-examples/src/test/java/org/
neo4j/examples/PathFindingDocTest.java>
Calculating the shortest path (least number of relationships) between two nodes:
Node startNode = graphDb.createNode();
Node middleNode1 = graphDb.createNode();
Node endNode = graphDb.createNode();
createRelationshipsBetween( startNode, middleNode1, endNode );
createRelationshipsBetween( startNode, middleNode2, middleNode3, endNode );
// Will find the shortest path between startNode and endNode via
// "MY_TYPE" relationships (in OUTGOING direction), like f.ex:
//
// (startNode)-->(middleNode1)-->(endNode)
//
PathFinder<Path> finder = GraphAlgoFactory.shortestPath(
Traversal.expanderForTypes( ExampleTypes.MY_TYPE, Direction.OUTGOING ), 15 );
Iterable<Path> paths = finder.findAllPaths( startNode, endNode );
Using Dijkstras algorithm <http://en.wikipedia.org/wiki/Dijkstra%27s_algorithm> to calculate

cheapest path between node A and B where each relationship can have a weight (i.e. cost) and the
path(s) with least cost are found.
PathFinder<WeightedPath> finder = GraphAlgoFactory.dijkstra(
Traversal.expanderForTypes( ExampleTypes.MY_TYPE, Direction.BOTH ), "cost" );
WeightedPath path = finder.findSinglePath( nodeA, nodeB );
// Get the weight for the found path
path.weight();
Using A* <http://en.wikipedia.org/wiki/A*_search_algorithm> to calculate the cheapest path between

node A and B, where cheapest is for example the path in a network of roads which has the shortest
length between node A and B. Heres our example graph:
Node nodeA =
Node nodeB =
Node nodeC =
Relationship
Relationship
Relationship
createNode( "name", "A", "x", 0d, "y", 0d

createNode( "name", "B", "x", 7d, "y", 0d
createNode( "name", "C", "x", 2d, "y", 1d
relAB = createRelationship( nodeA, nodeC,
relBC = createRelationship( nodeC, nodeB,
relAC = createRelationship( nodeA, nodeB,
);
);
);
"length", 2d );
"length", 3d );
"length", 10d );
EstimateEvaluator<Double> estimateEvaluator = new EstimateEvaluator<Double>()
536

{
@Override
public Double getCost( final Node node, final Node
{
double dx = (Double) node.getProperty( "x" ) double dy = (Double) node.getProperty( "y" ) double result = Math.sqrt( Math.pow( dx, 2 ) +
return result;
}
goal )
(Double) goal.getProperty( "x" );
(Double) goal.getProperty( "y" );
Math.pow( dy, 2 ) );
};
PathFinder<WeightedPath> astar = GraphAlgoFactory.aStar(
Traversal.expanderForAllTypes(),
CommonEvaluators.doubleCostEvaluator( "length" ), estimateEvaluator );
WeightedPath path = astar.findSinglePath( nodeA, nodeB );
537
32.10.Reading a management attribute

The JmxUtils <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/jmx/JmxUtils.html> class
includes methods to access Neo4j management beans. The common JMX service can be used as well,
but from your code you probably rather want to use the approach outlined here.
Tip
The source code of the example is found here: JmxDocTest.java <https://github.com/

neo4j/neo4j/blob/2.0.0/community/embedded-examples/src/test/java/org/neo4j/examples/
JmxDocTest.java>
This example shows how to get the start time of a database:
private static Date getStartTimeFromManagementBean(
GraphDatabaseService graphDbService )
{
GraphDatabaseAPI graphDb = (GraphDatabaseAPI) graphDbService;
ObjectName objectName = JmxUtils.getObjectName( graphDb, "Kernel" );
Date date = JmxUtils.getAttribute( objectName, "KernelStartTime" );
return date;
}
Depending on which Neo4j edition you are using different sets of management beans are available.
For all editions, see the org.neo4j.jmx <http://components.neo4j.org/neo4j-jmx/2.0.0/apidocs/org/
neo4j/jmx/package-summary.html> package.
For the Enterprise edition, see the org.neo4j.management <http://components.neo4j.org/neo4jmanagement/2.0.0/apidocs/org/neo4j/management/package-summary.html> package as well.
538
32.11.How to create unique nodes

This section is about how to ensure uniqueness of a property when creating nodes. For an overview of
the topic, see Section16.6, Creating unique nodes.
32.11.1.Get or create unique node using Cypher and unique constraints

Create a Cypher execution engine and a unique constraint.
try ( Transaction tx = graphdb.beginTx() )
{
graphdb.schema()
.constraintFor( DynamicLabel.label( "User" ) )
.assertPropertyIsUnique( "name" )
.create();
tx.success();
}
return new ExecutionEngine( graphdb() );
Use MERGE to create a unique node.

Node result = null;
ResourceIterator<Node> resultIterator = null;
{
String queryString = "MERGE (n:User {name: {name}}) RETURN n";
Map<String, Object> parameters = new HashMap<>();
parameters.put( "name", username );
resultIterator = engine.execute( queryString, parameters ).columnAs( "n" );
result = resultIterator.next();
tx.success();
return result;
}
32.11.2.Get or create unique node using a legacy index

Important
or create unique node using Cypher and unique constraints instead.
By using put-if-absent <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/index/
Index.html#putIfAbsent%28T,%20java.lang.String,%20java.lang.Object%29> functionality, entity
uniqueness can be guaranteed using an index.
Here the index acts as the lock and will only lock the smallest part needed to guarantee uniqueness
across threads and transactions. To get the more high-level get-or-create functionality make use
of UniqueFactory <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/index/
UniqueFactory.html> as seen in the example below.
Create a factory for unique nodes at application start.
{
UniqueFactory.UniqueNodeFactory result = new UniqueFactory.UniqueNodeFactory( graphDb, "users" )
{
@Override
protected void initialize( Node created, Map<String, Object> properties )
{
created.addLabel( DynamicLabel.label( "User" ) );
created.setProperty( "name", properties.get( "name" ) );
539

}
};
tx.success();
return result;
}
Use the unique node factory to get or create a node.

{
Node node = factory.getOrCreate( "name", username );
tx.success();
return node;
}
32.11.3.Pessimistic locking for node creation

Important
or create unique node using Cypher and unique constraints instead.
One might be tempted to use Java synchronization for pessimistic locking, but this is dangerous. By
mixing locks in Neo4j and in the Java runtime, it is easy to produce deadlocks that are not detectable
by Neo4j. As long as all locking is done by Neo4j, all deadlocks will be detected and avoided. Also, a
solution using manual synchronization doesnt ensure uniqueness in an HA environment.
This example uses a single lock node for locking. We create it only as a place to put locks, nothing
else.
Create a lock node at application start.
{
final Node lockNode = graphDb.createNode();
tx.success();
return lockNode;
}
Use the lock node to ensure nodes are not created concurrently.
{
Index<Node> usersIndex = graphDb.index().forNodes( "users" );
Node userNode = usersIndex.get( "name", username ).getSingle();
if ( userNode != null )
{
return userNode;
}
tx.acquireWriteLock( lockNode );
userNode = usersIndex.get( "name", username ).getSingle();
if ( userNode == null )
{
userNode = graphDb.createNode( DynamicLabel.label( "User" ) );
usersIndex.add( userNode, "name", username );
userNode.setProperty( "name", username );
}
tx.success();
return userNode;
}
Note that finishing the transaction will release the lock on the lock node.
540
32.12.Execute Cypher Queries from Java

Tip
The full source code of the example: JavaQuery.java <https://github.com/neo4j/neo4j/

blob/2.0.0/community/cypher/cypher-docs/src/test/java/org/neo4j/cypher/javacompat/
JavaQuery.java>
In Java, you can use the Cypher query language as per the example below. First, lets add some data.
GraphDatabaseService db = new GraphDatabaseFactory().newEmbeddedDatabase( DB_PATH );
try ( Transaction tx = db.beginTx(); )
{
Node myNode = db.createNode();
myNode.setProperty( "name", "my node" );
tx.success();
}
Execute a query:
ExecutionEngine engine = new ExecutionEngine( db );
ExecutionResult result;
try ( Transaction ignored = db.beginTx() )
{
result = engine.execute( "start n=node(*) where n.name = 'my node' return n, n.name" );
Note
Keep the ExecutionEngine around, dont create a new one for each query!
The result will be:
+-------------------------------------+
| n
| n.name
|
+-------------------------------------+
| Node[0]{name:"my node"} | "my node" |
+-------------------------------------+
1 row
Caution
The classes used here are from the org.neo4j.cypher.javacompat package, not
org.neo4j.cypher, see link to the Java API below.
You can get a list of the columns in the result:
List<String> columns = result.columns();
This contains:
[n, n.name]
To fetch the result items from a single column, do like this:

Iterator<Node> n_column = result.columnAs( "n" );
for ( Node node : IteratorUtil.asIterable( n_column ) )
{
// note: we're grabbing the name property from the node,
// not from the n.name in this case.
nodeResult = node + ": " + node.getProperty( "name" );
}
541

In this case theres only one node in the result:
Node[0]: my node
To get all columns, do like this instead:

for ( Map<String, Object> row : result )
{
for ( Entry<String, Object> column : row.entrySet() )
{
rows += column.getKey() + ": " + column.getValue() + "; ";
}
rows += "\n";
}
This outputs:
n.name: my node; n: Node[0];
Caution
dumpToString(), columnAs()
and iterator() cannot be called more than once on the same

ExecutionResult object. You should instead use only one and if you need the facilities of
the other methods on the same query result instead create a new ExecutionResult.
Caution
When using an ExecutionResult, youll need to exhaust it by using any of the iterating
methods (columnAs() and iterator()) on it. Failing to do so will not properly clean up
resources used by the ExecutionResult, leading to unwanted behavior, such as leaking
transactions.
For more information on the Java interface to Cypher, see the Java API <http://components.neo4j.org/
neo4j-cypher/2.0.0/apidocs/index.html>.
For more information and examples for Cypher, see PartIII, Cypher Query Language and
Chapter5, Data Modeling Examples.
542
32.13.Query Parameters
For more information on parameters see Section7.5, Parameters.
Below follows example of how to use parameters when executing Cypher queries from Java.
Node id.
Map<String, Object> params = new HashMap<String, Object>();
params.put( "id", 0 );
String query = "START n=node({id}) RETURN n.name";
ExecutionResult result = engine.execute( query, params );
Node object.
params.put( "node", andreasNode );
String query = "START n=node({node}) RETURN n.name";
Multiple node ids.

params.put( "id", Arrays.asList( 0, 1, 2 ) );
String query = "START n=node({id}) RETURN n.name";
String literal.
params.put( "name", "Johan" );
String query = "MATCH (n) WHERE n.name = {name} RETURN n";
Index value.
params.put( "value", "Michaela" );
String query = "START n=node:people(name = {value}) RETURN n";
Index query.
params.put( "query", "name:Andreas" );
String query = "START n=node:people({query}) RETURN n";
Numeric parameters for SKIP and LIMIT.

params.put( "s", 1 );
params.put( "l", 1 );
String query = "MATCH (n) RETURN n.name SKIP {s} LIMIT {l}";
Regular expression.
params.put( "regex", ".*h.*" );
String query = "MATCH (n) WHERE n.name =~ {regex} RETURN n.name";
Create node with properties.

Map<String, Object> props = new HashMap<String, Object>();
props.put( "name", "Andres" );
543

props.put( "position", "Developer" );
params.put( "props", props );
String query = "CREATE ({props})";
engine.execute( query, params );
Create multiple nodes with properties.

Map<String, Object> n1 = new HashMap<String, Object>();
n1.put( "name", "Andres" );
n1.put( "position", "Developer" );
n1.put( "awesome", true );
Map<String, Object> n2 = new HashMap<String, Object>();
n2.put( "name", "Michael" );
n2.put( "children", 3 );
List<Map<String, Object>> maps = Arrays.asList( n1, n2 );
params.put( "props", maps );
String query = "CREATE (n:Person {props}) RETURN n";
Setting all properties on node.

Map<String, Object> n1 = new HashMap<>();
n1.put( "name", "Andres" );
Map<String, Object> params = new HashMap<>();
params.put( "props", n1 );
String query = "MATCH (n) WHERE n.name='Michaela' SET n = {props}";
544
Chapter33.The Traversal Framework

The Neo4j Traversal API <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/
traversal/package-summary.html> is a callback based, lazily executed way of specifying desired
movements through a graph in Java. Some traversal examples are collected under Section32.7,
Traversal.
You can also use The Cypher Query Language as a powerful declarative way to query the graph.
545
The Traversal Framework
33.1.Main concepts
Here follows a short explanation of all different methods that can modify or add to a traversal
description.
Expanders define what to traverse, typically in terms of relationship direction and type.
Order for example depth-first or breadth-first.
Uniqueness visit nodes (relationships, paths) only once.
Evaluator decide what to return and whether to stop or continue traversal beyond the current
position.
Starting nodes where the traversal will begin.
Dept h First
Breadt h First
Order
Include/Exclude
Unique Relat ionships
Unique Pat hs
Evaluat or
None
Uniqueness
where t o go next
ret urn and prune policy avoid duplicat es
Prune/Cont inue
Traversal Descript ion
st art ing point of
Traverser
result as result as
Pat hs
Nodes
result as
Relat ionships
See Section33.2, Traversal Framework Java API for more details.
546
Direct ion
Expander
what t o t raverse
Relat ionship Type
applies
A Node
Unique Nodes
33.2.Traversal Framework Java API

The traversal framework consists of a few main interfaces in addition to Node and Relationship:
TraversalDescription, Evaluator, Traverser and Uniqueness are the main ones. The Path interface also
has a special purpose in traversals, since it is used to represent a position in the graph when evaluating
that position. Furthermore the PathExpander (replacing RelationshipExpander) and Expander interfaces
are central to traversals, but users of the API rarely need to implement them. There are also a set of
interfaces for advanced use, when explicit control over the traversal order is required: BranchSelector,
BranchOrderingPolicy and TraversalBranch.
33.2.1.TraversalDescription
The TraversalDescription <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/
traversal/TraversalDescription.html> is the main interface used for defining and initializing traversals.
It is not meant to be implemented by users of the traversal framework, but rather to be provided
by the implementation of the traversal framework as a way for the user to describe traversals.
TraversalDescription instances are immutable and its methods returns a new TraversalDescription that
is modified compared to the object the method was invoked on with the arguments of the method.
Relationships
Adds a relationship type to the list of relationship types to traverse. By default that list is
empty and it means that it will traverse all relationships, irregardless of type. If one or more
relationships are added to this list only the added types will be traversed. There are two
methods, one including direction <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/
neo4j/graphdb/traversal/TraversalDescription.html#relationships> and another one excluding
direction <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/traversal/
TraversalDescription.html#relationships>, where the latter traverses relationships in both directions
<http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/Direction.html#BOTH>.
33.2.2.Evaluator
<http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/traversal/
Evaluator.html>s are used for deciding, at each position (represented as a Path): should the traversal
continue, and/or should the node be included in the result. Given a Path, it asks for one of four actions
for that branch of the traversal:
Evaluator
Evaluation.INCLUDE_AND_CONTINUE:
Include this node in the result and continue the traversal

Evaluation.INCLUDE_AND_PRUNE: Include this node in the result, but dont continue the traversal
Evaluation.EXCLUDE_AND_CONTINUE: Exclude this node from the result, but continue the traversal
Evaluation.EXCLUDE_AND_PRUNE: Exclude this node from the result and dont continue the traversal
More than one evaluator can be added. Note that evaluators will be called for all positions the
traverser encounters, even for the start node.
33.2.3.Traverser
The Traverser <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/
traversal/Traverser.html> object is the result of invoking traverse() <http://
components.neo4j.org/neo4j-kernel/2.0.0/apidocs/org/neo4j/graphdb/traversal/
TraversalDescription.html#traverse(org.neo4j.graphdb.Node)> of a TraversalDescription object. It
represents a traversal positioned in the graph, and a specification of the format of the result. The actual
traversal is performed lazily each time the next()-method of the iterator of the Traverser is invoked.
547
33.2.4.Uniqueness
Sets the rules for how positions can be revisited during a traversal as stated in Uniqueness
<http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/kernel/Uniqueness.html>. Default
if not set is NODE_GLOBAL <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/kernel/
Uniqueness.html#NODE_GLOBAL>.
A Uniqueness can be supplied to the TraversalDescription to dictate under what circumstances a
traversal may revisit the same position in the graph. The various uniqueness levels that can be used in
Neo4j are:
NONE: Any position in the graph may be revisited.
NODE_GLOBAL uniqueness: No node in the entire graph may be visited more than once. This could
potentially consume a lot of memory since it requires keeping an in-memory data structure
remembering all the visited nodes.
RELATIONSHIP_GLOBAL uniqueness: no relationship in the entire graph may be visited more than once.
For the same reasons as NODE_GLOBAL uniqueness, this could use up a lot of memory. But since
graphs typically have a larger number of relationships than nodes, the memory overhead of this
uniqueness level could grow even quicker.
NODE_PATH uniqueness: A node may not occur previously in the path reaching up to it.
RELATIONSHIP_PATH uniqueness: A relationship may not occur previously in the path reaching up to it.
NODE_RECENT uniqueness: Similar to NODE_GLOBAL uniqueness in that there is a global collection of
visited nodes each position is checked against. This uniqueness level does however have a cap on
how much memory it may consume in the form of a collection that only contains the most recently
visited nodes. The size of this collection can be specified by providing a number as the second
argument to the TraversalDescription.uniqueness()-method along with the uniqueness level.
RELATIONSHIP_RECENT uniqueness: Works like NODE_RECENT uniqueness, but with relationships instead
of nodes.
Depth First / Breadth First
These are convenience methods for setting preorder depth-first <http://en.wikipedia.org/
wiki/Depth-first_search>/ breadth-first <http://en.wikipedia.org/wiki/Breadth-first_search>
BranchSelector|ordering policies. The same result can be achieved by calling the
order <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/traversal/
TraversalDescription.html#order> method with ordering policies from the Traversal <http://
components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/kernel/Traversal.html#preorderDepthFirst>
factory <http://components.neo4j.org/neo4j-kernel/2.0.0/apidocs/org/neo4j/kernel/
Traversal.html#preorderBreadthFirst>, or to write your own BranchSelector/BranchOrderingPolicy and
pass in.
33.2.5.Order How to move through branches?

A more generic version of depthFirst/breadthFirst methods in that it allows an arbitrary
BranchOrderingPolicy <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/traversal/
BranchOrderingPolicy.html> to be injected into the description.
33.2.6.BranchSelector
A BranchSelector is used for selecting which branch of the traversal to attempt next. This is used
for implementing traversal orderings. The traversal framework provides a few basic ordering
implementations:
548

Traversal.preorderDepthFirst(): Traversing depth first, visiting each node before visiting its child
nodes.
Traversal.postorderDepthFirst(): Traversing depth first, visiting each node after visiting its child
nodes.
Traversal.preorderBreadthFirst(): Traversing breadth first, visiting each node before visiting its
child nodes.
Traversal.postorderBreadthFirst(): Traversing breadth first, visiting each node after visiting its
child nodes.
Note
Please note that breadth first traversals have a higher memory overhead than depth first
traversals.
BranchSelectors carries state and hence needs to be uniquely instantiated for each traversal. Therefore
it is supplied to the TraversalDescription through a BranchOrderingPolicy interface, which is a factory
of BranchSelector instances.
A user of the Traversal framework rarely needs to implement his own BranchSelector or
BranchOrderingPolicy, it is provided to let graph algorithm implementors provide their own traversal
orders. The Neo4j Graph Algorithms package contains for example a BestFirst order BranchSelector/
BranchOrderingPolicy that is used in BestFirst search algorithms such as A* and Dijkstra.
BranchOrderingPolicy
A factory for creating BranchSelectors to decide in what order branches are returned (where
a branchs position is represented as a Path <http://components.neo4j.org/neo4j/2.0.0/
apidocs/org/neo4j/graphdb/Path.html> from the start node to the current node). Common
policies are depth-first <http://components.neo4j.org/neo4j-kernel/2.0.0/apidocs/org/
neo4j/graphdb/traversal/TraversalDescription.html#depthFirst()> and breadth-first
<http://components.neo4j.org/neo4j-kernel/2.0.0/apidocs/org/neo4j/graphdb/traversal/
TraversalDescription.html#breadthFirst()> and thats why there are convenience methods for those.
For example, calling TraversalDescriptiondepthFirst() <http://components.neo4j.org/neo4j/2.0.0/
apidocs/org/neo4j/graphdb/traversal/TraversalDescription.html#depthFirst()> is equivalent to:
description.order( Traversal.preorderDepthFirst() );
TraversalBranch
An object used by the BranchSelector to get more branches from a certain branch. In essence these are
a composite of a Path and a RelationshipExpander that can be used to get new TraversalBranch <http://
components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/traversal/TraversalBranch.html>es from
the current one.
33.2.7.Path
A Path <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/Path.html> is a general
interface that is part of the Neo4j API. In the traversal API of Neo4j the use of Paths are twofold.
Traversers can return their results in the form of the Paths of the visited positions in the graph that are
marked for being returned. Path objects are also used in the evaluation of positions in the graph, for
determining if the traversal should continue from a certain point or not, and whether a certain position
should be included in the result set or not.
549
33.2.8.PathExpander/RelationshipExpander
The traversal framework use PathExpanders (replacing RelationshipExpander) to discover the
relationships that should be followed from a particular path to further branches in the traversal.
33.2.9.Expander
A more generic version of relationships where a RelationshipExpander is injected, defining all
relationships to be traversed for any given node. By default (and when using relationships)
a default expander <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/kernel/
Traversal.html#emptyExpander> is used, where any particular order of relationships isnt
guaranteed. Theres another implementation which guarantees that relationships are traversed in
order of relationship type <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/kernel/
OrderedByTypeExpander.html>, where types are iterated in the order they were added.
The Expander interface is an extension of the RelationshipExpander interface that makes it possible to
build customized versions of an Expander. The implementation of TraversalDescription uses this to
provide methods for defining which relationship types to traverse, this is the usual way a user of the
API would define a RelationshipExpander by building it internally in the TraversalDescription.
All the RelationshipExpanders provided by the Neo4j traversal framework also implement the
Expander interface. For a user of the traversal API it is easier to implement the PathExpander/
RelationshipExpander interface, since it only contains one method the method for getting the
relationships from a path/node, the methods that the Expander interface adds are just for building new
Expanders.
33.2.10.How to use the Traversal framework

A traversal description <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/
traversal/TraversalDescription.html> is built using a fluent interface and such a description can then
spawn traversers <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/traversal/
Traverser.html>.
550

Figure33.1.Traversal Example Graph
Node[ 3] nam e = 'Lisa'
LIKES
Node[ 2] nam e = 'Ed'
KNOWS
KNOWS
Node[ 0] nam e = 'Lars'
KNOWS
Node[ 5] nam e = 'Dirk'
KNOWS
KNOWS
KNOWS
With the definition of the RelationshipTypes as

private enum Rels implements RelationshipType
{
LIKES, KNOWS
}
The graph can be traversed with for example the following traverser, starting at the Joe node:
for ( Path position : Traversal.description()
.depthFirst()
.relationships( Rels.KNOWS )
.relationships( Rels.LIKES, Direction.INCOMING )
.evaluator( Evaluators.toDepth( 5 ) )
.traverse( node ) )
{
output += position + "\n";
}
The traversal will output:

(6)
(6)<--[LIKES,1]--(3)
(6)<--[LIKES,1]--(3)--[KNOWS,6]-->(0)
(6)<--[LIKES,1]--(3)--[KNOWS,6]-->(0)--[KNOWS,4]-->(5)
(6)<--[LIKES,1]--(3)--[KNOWS,6]-->(0)--[KNOWS,4]-->(5)--[KNOWS,3]-->(4)
551

(6)<--[LIKES,1]--(3)--[KNOWS,6]-->(0)--[KNOWS,4]-->(5)--[KNOWS,3]-->(4)--[KNOWS,2]-->(1)
(6)<--[LIKES,1]--(3)--[KNOWS,6]-->(0)<--[KNOWS,5]--(2)
Since TraversalDescription <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/

traversal/TraversalDescription.html>s are immutable it is also useful to create template descriptions
which holds common settings shared by different traversals. For example, lets start with this
traverser:
final TraversalDescription FRIENDS_TRAVERSAL = Traversal.description()
.depthFirst()
.relationships( Rels.KNOWS )
.uniqueness( Uniqueness.RELATIONSHIP_GLOBAL );
This traverser would yield the following output (we will keep starting from the Joe node):
(6)
(6)--[KNOWS,0]-->(1)
(6)--[KNOWS,0]-->(1)<--[KNOWS,2]--(4)
(6)--[KNOWS,0]-->(1)<--[KNOWS,2]--(4)<--[KNOWS,3]--(5)
(6)--[KNOWS,0]-->(1)<--[KNOWS,2]--(4)<--[KNOWS,3]--(5)<--[KNOWS,4]--(0)
(6)--[KNOWS,0]-->(1)<--[KNOWS,2]--(4)<--[KNOWS,3]--(5)<--[KNOWS,4]--(0)<--[KNOWS,5]--(2)
(6)--[KNOWS,0]-->(1)<--[KNOWS,2]--(4)<--[KNOWS,3]--(5)<--[KNOWS,4]--(0)<--[KNOWS,6]--(3)
Now lets create a new traverser from it, restricting depth to three:
for ( Path path : FRIENDS_TRAVERSAL
.traverse( node ) )
{
output += path + "\n";
}
This will give us the following result:

(6)
(6)--[KNOWS,0]-->(1)
(6)--[KNOWS,0]-->(1)<--[KNOWS,2]--(4)
Or how about from depth two to four? Thats done like this:
for ( Path path : FRIENDS_TRAVERSAL
.evaluator( Evaluators.fromDepth( 2 ) )
.traverse( node ) )
{
output += path + "\n";
}
This traversal gives us:

(6)--[KNOWS,0]-->(1)<--[KNOWS,2]--(4)
(6)--[KNOWS,0]-->(1)<--[KNOWS,2]--(4)<--[KNOWS,3]--(5)<--[KNOWS,4]--(0)
For various useful evaluators, see the Evaluators <http://components.neo4j.org/neo4j/2.0.0/apidocs/

org/neo4j/graphdb/traversal/Evaluators.html> Java API or simply implement the Evaluator <http://
components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/traversal/Evaluator.html> interface
yourself.
If youre not interested in the Path <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/
graphdb/Path.html>s, but the Node <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/
neo4j/graphdb/Node.html>s you can transform the traverser into an iterable of nodes <http://
552

components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/traversal/Traverser.html#nodes()> like
this:
for ( Node currentNode : FRIENDS_TRAVERSAL
.traverse( node )
.nodes() )
{
output += currentNode.getProperty( "name" ) + "\n";
}
In this case we use it to retrieve the names:

Joe
Sara
Peter
Dirk
Lars
Ed
Lisa
Relationships <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/traversal/
Traverser.html#relationships()> are fine as well, heres how to get them:
for ( Relationship relationship : FRIENDS_TRAVERSAL
.traverse( node )
.relationships() )
{
output += relationship.getType() + "\n";
}
Here the relationship types are written, and we get:

RelationshipTypeToken[name:KNOWS,
id:0]
id:0]
id:0]
id:0]
id:0]
id:0]
Tip
The source code for the traversers in this example is available at: TraversalExample.java
<https://github.com/neo4j/neo4j/blob/2.0.0/community/embedded-examples/src/main/java/
org/neo4j/examples/TraversalExample.java>
553
Chapter34.Legacy Indexing
Note
This is not the same as indexes defined in the schema, the documentation below is for the
legacy indexing in Neo4j.
This chapter focuses on how to use the Manual Indexes and Autoindexes. As of Neo4j 2.0, this is not
the favored method of indexing data in Neo4j, instead we recommend defining indexes in the database
schema.
However, support for legacy indexes remains, because certain features, such as uniqueness
constraints, are not yet handled by the new indexes.
554
Legacy Indexing
34.1.Introduction
Legacy Indexing operations are part of the Neo4j index API <http://components.neo4j.org/
neo4j/2.0.0/apidocs/org/neo4j/graphdb/index/package-summary.html>.
Each index is tied to a unique, user-specified name (for example "first_name" or "books") and
can index either nodes <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/
Node.html> or relationships <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/
Relationship.html>.
The default index implementation is provided by the neo4j-lucene-index component, which is included
in the standard Neo4j download. It can also be downloaded separately from http://repo1.maven.org/
maven2/org/neo4j/neo4j-lucene-index/ . For Maven users, the neo4j-lucene-index component
has the coordinates org.neo4j:neo4j-lucene-index and should be used with the same version of
org.neo4j:neo4j-kernel. Different versions of the index and kernel components are not compatible in
the general case. Both components are included transitively by the org.neo4j:neo4j:pom artifact which
makes it simple to keep the versions in sync.
For initial import of data using indexes, see Section35.3, Index Batch Insertion.
Note
All modifying index operations must be performed inside a transaction, as with any
modifying operation in Neo4j.
555
Legacy Indexing
34.2.Create
An index is created if it doesnt exist when you ask for it. Unless you give it a custom configuration, it
will be created with default configuration and backend.
To set the stage for our examples, lets create some indexes to begin with:
IndexManager index = graphDb.index();
Index<Node> actors = index.forNodes( "actors" );
Index<Node> movies = index.forNodes( "movies" );
RelationshipIndex roles = index.forRelationships( "roles" );
This will create two node indexes and one relationship index with default configuration. See
Section34.8, Relationship indexes for more information specific to relationship indexes.
See Section34.10, Configuration and fulltext indexes for how to create fulltext indexes.
You can also check if an index exists like this:
boolean indexExists = index.existsForNodes( "actors" );
556
Legacy Indexing
34.3.Delete
Indexes can be deleted. When deleting, the entire contents of the index will be removed as well as its
associated configuration. An index can be created with the same name at a later point in time.
Index<Node> actors = index.forNodes( "actors" );
actors.delete();
Note that the actual deletion of the index is made during the commit of the surrounding transaction.
Calls made to such an index instance after delete() <http://components.neo4j.org/neo4j/2.0.0/
apidocs/org/neo4j/graphdb/index/Index.html#delete%28%29> has been called are invalid inside that
transaction as well as outside (if the transaction is successful), but will become valid again if the
transaction is rolled back.
557
Legacy Indexing
34.4.Add
Each index supports associating any number of key-value pairs with any number of entities (nodes or
relationships), where each association between entity and key-value pair is performed individually. To
begin with, lets add a few nodes to the indexes:
// Actors
Node reeves = graphDb.createNode();
reeves.setProperty( "name", "Keanu Reeves" );
actors.add( reeves, "name", reeves.getProperty( "name" ) );
Node bellucci = graphDb.createNode();
bellucci.setProperty( "name", "Monica Bellucci" );
actors.add( bellucci, "name", bellucci.getProperty( "name" ) );
// multiple values for a field, in this case for search only
// and not stored as a property.
actors.add( bellucci, "name", "La Bellucci" );
// Movies
Node theMatrix = graphDb.createNode();
theMatrix.setProperty( "title", "The Matrix" );
theMatrix.setProperty( "year", 1999 );
movies.add( theMatrix, "title", theMatrix.getProperty( "title" ) );
movies.add( theMatrix, "year", theMatrix.getProperty( "year" ) );
Node theMatrixReloaded = graphDb.createNode();
theMatrixReloaded.setProperty( "title", "The Matrix Reloaded" );
theMatrixReloaded.setProperty( "year", 2003 );
movies.add( theMatrixReloaded, "title", theMatrixReloaded.getProperty( "title" ) );
movies.add( theMatrixReloaded, "year", 2003 );
Node malena = graphDb.createNode();
malena.setProperty( "title", "Malna" );
malena.setProperty( "year", 2000 );
movies.add( malena, "title", malena.getProperty( "title" ) );
movies.add( malena, "year", malena.getProperty( "year" ) );
Note that there can be multiple values associated with the same entity and key.
Next up, well create relationships and index them as well:
// we need a relationship type
DynamicRelationshipType ACTS_IN = DynamicRelationshipType.withName( "ACTS_IN" );
// create relationships
Relationship role1 = reeves.createRelationshipTo( theMatrix, ACTS_IN );
role1.setProperty( "name", "Neo" );
roles.add( role1, "name", role1.getProperty( "name" ) );
Relationship role2 = reeves.createRelationshipTo( theMatrixReloaded, ACTS_IN );
role2.setProperty( "name", "Neo" );
Relationship role3 = bellucci.createRelationshipTo( theMatrixReloaded, ACTS_IN );
role3.setProperty( "name", "Persephone" );
Relationship role4 = bellucci.createRelationshipTo( malena, ACTS_IN );
role4.setProperty( "name", "Malna Scordia" );
After these operations, our example graph looks like this:
558
Legacy Indexing
Figure34.1.Movie and Actor Graph
nam e = 'Keanu Reeves'
ACTS_IN
nam e = 'Neo'
t it le = 'The Mat rix'

year = 1999
nam e = 'Monica Bellucci'
ACTS_IN
nam e = 'Neo'
ACTS_IN
nam e = 'Persephone'
t it le = 'The Mat rix Reloaded'

year = 2003
559
ACTS_IN
nam e = 'Malna Scordia'
t it le = 'Malna'
year = 2000
Legacy Indexing
34.5.Remove
Removing <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/index/
Index.html#remove%28T,%20java.lang.String,%20java.lang.Object%29> from an index is similar to
adding, but can be done by supplying one of the following combinations of arguments:
entity
entity, key
entity, key, value
// completely remove bellucci from the actors index
actors.remove( bellucci );
// remove any "name" entry of bellucci from the actors index
actors.remove( bellucci, "name" );
// remove the "name" -> "La Bellucci" entry of bellucci
actors.remove( bellucci, "name", "La Bellucci" );
560
Legacy Indexing
34.6.Update
Important
To update an index entry, the old one must be removed and a new one added. For details
on removing index entries, see Section34.5, Remove.
Remember that a node or relationship can be associated with any number of key-value pairs in an
index. This means that you can index a node or relationship with many key-value pairs that have
the same key. In the case where a property value changes and youd like to update the index, its not
enough to just index the new value youll have to remove the old value as well.
Heres a code example that demonstrates how its done:
// create a node with a property
// so we have something to update later on
Node fishburn = graphDb.createNode();
fishburn.setProperty( "name", "Fishburn" );
// index it
actors.add( fishburn, "name", fishburn.getProperty( "name" ) );
// update the index entry
// when the property value changes
actors.remove( fishburn, "name", fishburn.getProperty( "name" ) );
fishburn.setProperty( "name", "Laurence Fishburn" );
actors.add( fishburn, "name", fishburn.getProperty( "name" ) );
561
Legacy Indexing
34.7.Search
An index can be searched in two ways, get <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/
neo4j/graphdb/index/Index.html#get%28java.lang.String,%20java.lang.Object%29> and query
<http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/index/Index.html#query
%28java.lang.String,%20java.lang.Object%29>. The get method will return exact matches to
the given key-value pair, whereas query exposes querying capabilities directly from the backend
used by the index. For example the Lucene query syntax <http://lucene.apache.org/core/3_6_2/
queryparsersyntax.html> can be used directly with the default indexing backend.
34.7.1.Get
This is how to search for a single exact match:
IndexHits<Node> hits = actors.get( "name", "Keanu Reeves" );
Node reeves = hits.getSingle();
IndexHits <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/index/
IndexHits.html> is an Iterable with some additional useful methods. For example getSingle() <http://
components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/index/IndexHits.html#getSingle
%28%29> returns the first and only item from the result iterator, or null if there isnt any hit.
Heres how to get a single relationship by exact matching and retrieve its start and end nodes:
Relationship persephone = roles.get( "name", "Persephone" ).getSingle();
Node actor = persephone.getStartNode();
Node movie = persephone.getEndNode();
Finally, we can iterate over all exact matches from a relationship index:
for ( Relationship role : roles.get( "name", "Neo" ) )
{
// this will give us Reeves twice
Node reeves = role.getStartNode();
}
Important
In case you dont iterate through all the hits, IndexHits.close() <http://
components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/index/IndexHits.html#close
%28%29> must be called explicitly.
34.7.2.Query
There are two query methods, one which uses a key-value signature where the value represents a
query for values with the given key only. The other method is more generic and supports querying for
more than one key-value pair in the same query.
Heres an example using the key-query option:
for ( Node actor : actors.query( "name", "*e*" ) )
{
// This will return Reeves and Bellucci
}
In the following example the query uses multiple keys:

for ( Node movie : movies.query( "title:*Matrix* AND year:1999" ) )
{
// This will return "The Matrix" from 1999 only.
}
562
Legacy Indexing
Note
Beginning a wildcard search with "*" or "?" is discouraged by Lucene, but will
nevertheless work.
Caution
You cant have any whitespace in the search term with this syntax. See Section34.11.3,
Querying with Lucene Query objects for how to do that.
563
Legacy Indexing
34.8.Relationship indexes
An index for relationships is just like an index for nodes, extended by providing support to constrain
a search to relationships with a specific start and/or end nodes These extra methods reside in the
RelationshipIndex <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/graphdb/index/
RelationshipIndex.html> interface which extends Index<Relationship> <http://components.neo4j.org/
neo4j/2.0.0/apidocs/org/neo4j/graphdb/index/Index.html>.
Example of querying a relationship index:
// find relationships filtering on start node
// using exact matches
IndexHits<Relationship> reevesAsNeoHits;
reevesAsNeoHits = roles.get( "name", "Neo", reeves, null );
Relationship reevesAsNeo = reevesAsNeoHits.iterator().next();
reevesAsNeoHits.close();
// find relationships filtering on end node
// using a query
IndexHits<Relationship> matrixNeoHits;
matrixNeoHits = roles.query( "name", "*eo", null, theMatrix );
Relationship matrixNeo = matrixNeoHits.iterator().next();
matrixNeoHits.close();
And heres an example for the special case of searching for a specific relationship type:
// find relationships filtering on end node
// using a relationship type.
// this is how to add it to the index:
roles.add( reevesAsNeo, "type", reevesAsNeo.getType().name() );
// Note that to use a compound query, we can't combine committed
// and uncommitted index entries, so we'll commit before querying:
tx.success();
tx.finish();
// and now we can search for it:
{
IndexHits<Relationship> typeHits = roles.query( "type:ACTS_IN AND name:Neo", null, theMatrix );
Relationship typeNeo = typeHits.iterator().next();
typeHits.close();
Such an index can be useful if your domain has nodes with a very large number of relationships
between them, since it reduces the search time for a relationship between two nodes. A good example
where this approach pays dividends is in time series data, where we have readings represented as a
relationship per occurrence.
564
Legacy Indexing
34.9.Scores
The IndexHits interface exposes scoring <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/
graphdb/index/IndexHits.html#currentScore%28%29> so that the index can communicate scores
for the hits. Note that the result is not sorted by the score unless you explicitly specify that. See
Section34.11.2, Sorting for how to sort by score.
IndexHits<Node> hits = movies.query( "title", "The*" );
for ( Node movie : hits )
{
System.out.println( movie.getProperty( "title" ) + " " + hits.currentScore() );
}
565
Legacy Indexing
34.10.Configuration and fulltext indexes

At the time of creation extra configuration can be specified to control the behavior of the index and
which backend to use. For example to create a Lucene fulltext index:
Index<Node> fulltextMovies = index.forNodes( "movies-fulltext",
MapUtil.stringMap( IndexManager.PROVIDER, "lucene", "type", "fulltext" ) );
fulltextMovies.add( theMatrix, "title", "The Matrix" );
fulltextMovies.add( theMatrixReloaded, "title", "The Matrix Reloaded" );
// search in the fulltext index
Node found = fulltextMovies.query( "title", "reloAdEd" ).getSingle();
Heres an example of how to create an exact index which is case-insensitive:

Index<Node> index = graphDb.index().forNodes( "exact-case-insensitive",
stringMap( "type", "exact", "to_lower_case", "true" ) );
Node node = graphDb.createNode();
index.add( node, "name", "Thomas Anderson" );
assertContains( index.query( "name", "\"Thomas Anderson\"" ), node );
assertContains( index.query( "name", "\"thoMas ANDerson\"" ), node );
Tip
In order to search for tokenized words, the query method has to be used. The get method
will only match the full string value, not the tokens.
The configuration of the index is persisted once the index has been created. The provider
configuration key is interpreted by Neo4j, but any other configuration is passed onto the backend
index (e.g. Lucene) to interpret.
Lucene indexing configuration parameters
Parameter
Possible values
Effect
type
exact, fulltext
exact
to_lower_case
true, false
This parameter goes together with type: fulltext

and converts values to lower case during both
additions and querying, making the index case
insensitive. Defaults to true.
analyzer
the full class name of

an Analyzer <http://
lucene.apache.org/
core/3_6_2/api/core/org/
apache/lucene/analysis/
Analyzer.html>
Overrides the type so that a custom analyzer

can be used. Note: to_lower_case still affects
lowercasing of string queries. If the custom
analyzer uppercases the indexed tokens, string
queries will not match as expected.
is the default and uses a Lucene

keyword analyzer <http://lucene.apache.org/
core/3_6_2/api/core/org/apache/lucene/analysis/
KeywordAnalyzer.html>. fulltext uses a whitespace tokenizer in its analyzer.
566
Legacy Indexing
34.11.Extra features for Lucene indexes

34.11.1.Numeric ranges
Lucene supports smart indexing of numbers, querying for ranges and sorting such results, and so does
its backend for Neo4j. To mark a value so that it is indexed as a numeric value, we can make use of
the ValueContext <http://components.neo4j.org/neo4j-lucene-index/2.0.0/apidocs/org/neo4j/index/
lucene/ValueContext.html> class, like this:
movies.add( theMatrix, "year-numeric", new ValueContext( 1999 ).indexNumeric() );
movies.add( theMatrixReloaded, "year-numeric", new ValueContext( 2003 ).indexNumeric() );
movies.add( malena, "year-numeric", new ValueContext( 2000 ).indexNumeric() );
int from = 1997;
int to = 1999;
hits = movies.query( QueryContext.numericRange( "year-numeric", from, to ) );
Note
The same type must be used for indexing and querying. That is, you cant index a value as
a Long and then query the index using an Integer.
By giving null as from/to argument, an open ended query is created. In the following example we are
doing that, and have added sorting to the query as well:
hits = movies.query(
QueryContext.numericRange( "year-numeric", from, null )
.sortNumeric( "year-numeric", false ) );
From/to in the ranges defaults to be inclusive, but you can change this behavior by using two extra
parameters:
movies.add( theMatrix, "score", new ValueContext( 8.7 ).indexNumeric() );
movies.add( theMatrixReloaded, "score", new ValueContext( 7.1 ).indexNumeric() );
movies.add( malena, "score", new ValueContext( 7.4 ).indexNumeric() );
// include 8.0, exclude 9.0
hits = movies.query( QueryContext.numericRange( "score", 8.0, 9.0, true, false ) );
34.11.2.Sorting
Lucene performs sorting very well, and that is also exposed in the index backend, through the
QueryContext <http://components.neo4j.org/neo4j-lucene-index/2.0.0/apidocs/org/neo4j/index/
lucene/QueryContext.html> class:
hits = movies.query( "title", new QueryContext( "*" ).sort( "title" ) );
for ( Node hit : hits )
{
// all movies with a title in the index, ordered by title
}
// or
hits = movies.query( new QueryContext( "title:*" ).sort( "year", "title" ) );
for ( Node hit : hits )
{
// all movies with a title in the index, ordered by year, then title
}
We sort the results by relevance (score) like this:

hits = movies.query( "title", new QueryContext( "The*" ).sortByScore() );
{
567
Legacy Indexing
// hits sorted by relevance (score)
}
34.11.3.Querying with Lucene Query objects

Instead of passing in Lucene query syntax queries, you can instantiate such queries programmatically
and pass in as argument, for example:
// a TermQuery will give exact matches
Node actor = actors.query( new TermQuery( new Term( "name", "Keanu Reeves" ) ) ).getSingle();
Note that the TermQuery <http://lucene.apache.org/core/3_6_2/api/core/org/apache/lucene/search/

TermQuery.html> is basically the same thing as using the get method on the index.
This is how to perform wildcard searches using Lucene Query Objects:
hits = movies.query( new WildcardQuery( new Term( "title", "The Matrix*" ) ) );
{
System.out.println( movie.getProperty( "title" ) );
}
Note that this allows for whitespace in the search string.
34.11.4.Compound queries
Lucene supports querying for multiple terms in the same query, like so:
hits = movies.query( "title:*Matrix* AND year:1999" );
Caution
Compound queries cant search across committed index entries and those who havent got
committed yet at the same time.
34.11.5.Default operator
The default operator (that is whether AND or OR is used in between different terms) in a query is OR.
Changing that behavior is also done via the QueryContext <http://components.neo4j.org/neo4j-luceneindex/2.0.0/apidocs/org/neo4j/index/lucene/QueryContext.html> class:
QueryContext query = new QueryContext( "title:*Matrix* year:1999" )
.defaultOperator( Operator.AND );
hits = movies.query( query );
34.11.6.Caching
If your index lookups becomes a performance bottle neck, caching can be enabled for certain keys in
certain indexes (key locations) to speed up get requests. The caching is implemented with an LRU
<http://en.wikipedia.org/wiki/Cache_algorithms#Least_Recently_Used> cache so that only the most
recently accessed results are cached (with "results" meaning a query result of a get request, not a
single entity). You can control the size of the cache (the maximum number of results) per index key.
Index<Node> index = graphDb.index().forNodes( "actors" );
((LuceneIndex<Node>) index).setCacheCapacity( "name", 300000 );
Caution
This setting is not persisted after shutting down the database. This means: set this value
after each startup of the database if you want to keep it.
568
Legacy Indexing
34.12.Automatic Indexing
Neo4j provides a single index for nodes and one for relationships in each database that automatically
follow property values as they are added, deleted and changed on database primitives. This
functionality is called auto indexing and is controlled both from the database configuration Map and
through its own API.
34.12.1.Configuration
By default Auto Indexing is off for both Nodes and Relationships. To configure this in the
neo4j.properties file, use the configuration keys node_auto_indexing and relationship_auto_indexing.
For embedded mode, use the configuration options GraphDatabaseSettings.node_auto_indexing and
GraphDatabaseSettings.relationship_auto_indexing. In both cases, set the value to true. This will
enable automatic indexing on startup. Just note that were not done yet, see below!
To actually auto index something, you have to set which properties should get indexed. You do
this by listing the property keys to index on. In the configuration file, use the node_keys_indexable
and relationship_keys_indexable configuration keys. When using embedded mode, use the
GraphDatabaseSettings.node_keys_indexable and GraphDatabaseSettings.relationship_keys_indexable
configuration keys. In all cases, the value should be a comma separated list of property keys to index
on.
When coding in Java, its done like this:
/*
* Creating the configuration, adding nodeProp1 and nodeProp2 as
* auto indexed properties for Nodes and relProp1 and relProp2 as
* auto indexed properties for Relationships. Only those will be
* indexed. We also have to enable auto indexing for both these
* primitives explicitly.
*/
GraphDatabaseService graphDb = new GraphDatabaseFactory().
newEmbeddedDatabaseBuilder( storeDirectory ).
setConfig( GraphDatabaseSettings.node_keys_indexable, "nodeProp1,nodeProp2" ).
setConfig( GraphDatabaseSettings.relationship_keys_indexable, "relProp1,relProp2" ).
setConfig( GraphDatabaseSettings.node_auto_indexing, "true" ).
setConfig( GraphDatabaseSettings.relationship_auto_indexing, "true" ).
newGraphDatabase();
Node node1 = null, node2 = null;
Relationship rel = null;
{
// Create the primitives
node1 = graphDb.createNode();
rel = node1.createRelationshipTo( node2,
DynamicRelationshipType.withName( "DYNAMIC" ) );
// Add indexable and non-indexable properties
node1.setProperty( "nodeProp1", "nodeProp1Value" );
node2.setProperty( "nodeProp2", "nodeProp2Value" );
node1.setProperty( "nonIndexed", "nodeProp2NonIndexedValue" );
rel.setProperty( "relProp1", "relProp1Value" );
rel.setProperty( "relPropNonIndexed", "relPropValueNonIndexed" );
// Make things persistent
tx.success();
}
569
Legacy Indexing
34.12.2.Search
The usefulness of the auto indexing functionality comes of course from the ability to actually query
the index and retrieve results. To that end, you can acquire a ReadableIndex object from the AutoIndexer
that exposes all the query and get methods of a full Index <http://components.neo4j.org/neo4j/2.0.0/
apidocs/org/neo4j/graphdb/index/Index.html> with exactly the same functionality. Continuing from
the previous example, accessing the index is done like this:
{
// Get the Node auto index
ReadableIndex<Node> autoNodeIndex = graphDb.index()
.getNodeAutoIndexer()
.getAutoIndex();
// node1 and node2 both had auto indexed properties, get them
assertEquals( node1,
autoNodeIndex.get( "nodeProp1", "nodeProp1Value" ).getSingle() );
autoNodeIndex.get( "nodeProp2", "nodeProp2Value" ).getSingle() );
// node2 also had a property that should be ignored.
assertFalse( autoNodeIndex.get( "nonIndexed",
"nodeProp2NonIndexedValue" ).hasNext() );
// Get the relationship auto index
ReadableIndex<Relationship> autoRelIndex = graphDb.index()
.getRelationshipAutoIndexer()
.getAutoIndex();
// One property was set for auto indexing
assertEquals( rel,
autoRelIndex.get( "relProp1", "relProp1Value" ).getSingle() );
// The rest should be ignored
assertFalse( autoRelIndex.get( "relPropNonIndexed",
"relPropValueNonIndexed" ).hasNext() );
}
34.12.3.Runtime Configuration
The same options that are available during database creation via the configuration can also be set
during runtime via the AutoIndexer API.
Gaining access to the AutoIndexer API and adding two Node and one Relationship properties to auto
index is done like so:
// Start without any configuration
newEmbeddedDatabase( storeDirectory );
// Get the Node AutoIndexer, set nodeProp1 and nodeProp2 as auto
// indexed.
AutoIndexer<Node> nodeAutoIndexer = graphDb.index()
.getNodeAutoIndexer();
nodeAutoIndexer.startAutoIndexingProperty( "nodeProp1" );
nodeAutoIndexer.startAutoIndexingProperty( "nodeProp2" );
// Get the Relationship AutoIndexer
AutoIndexer<Relationship> relAutoIndexer = graphDb.index()
.getRelationshipAutoIndexer();
relAutoIndexer.startAutoIndexingProperty( "relProp1" );
// None of the AutoIndexers are enabled so far. Do that now
nodeAutoIndexer.setEnabled( true );
relAutoIndexer.setEnabled( true );
570
Legacy Indexing
Note
Parameters to the AutoIndexers passed through the Configuration and settings made
through the API are cumulative. So you can set some beforehand known settings, do
runtime checks to augment the initial configuration and then enable the desired auto
indexers - the final configuration is the same regardless of the method used to reach it.
34.12.4.Updating the Automatic Index

Updates to the auto indexed properties happen of course automatically as you update them. Removal
of properties from the auto index happens for two reasons. One is that you actually removed the
property. The other is that you stopped autoindexing on a property. When the latter happens, any
primitive you touch and it has that property, it is removed from the auto index, regardless of any
operations on the property. When you start or stop auto indexing on a property, no auto update
operation happens currently. If you need to change the set of auto indexed properties and have them
re-indexed, you currently have to do this by hand. An example will illustrate the above better:
/*
* Creating the configuration
*/
newEmbeddedDatabaseBuilder( storeDirectory ).
setConfig( GraphDatabaseSettings.node_keys_indexable, "nodeProp1,nodeProp2" ).
setConfig( GraphDatabaseSettings.node_auto_indexing, "true" ).
newGraphDatabase();
Node node1 = null, node2 = null, node3 = null, node4 = null;
{
// Create the primitives
// Add indexable and non-indexable properties
node1.setProperty( "nodeProp1", "nodeProp1Value"
);
);
);
);

tx.success();
}
/*
* Here both nodes are indexed. To demonstrate removal, we stop
* autoindexing nodeProp1.
*/
AutoIndexer<Node> nodeAutoIndexer = graphDb.index().getNodeAutoIndexer();
nodeAutoIndexer.stopAutoIndexingProperty( "nodeProp1" );
{
/*
* nodeProp1 is no longer auto indexed. It will be
* removed regardless. Note that node3 will remain.
*/
node1.setProperty( "nodeProp1", "nodeProp1Value2" );
/*
* node2 will be auto updated
*/
571
Legacy Indexing
node2.setProperty( "nodeProp2", "nodeProp2Value2" );
/*
* remove node4 property nodeProp2 from index.
*/
node4.removeProperty( "nodeProp2" );
tx.success();
}
{
// Verify
ReadableIndex<Node> nodeAutoIndex = nodeAutoIndexer.getAutoIndex();
// node1 is completely gone
assertFalse( nodeAutoIndex.get( "nodeProp1", "nodeProp1Value" ).hasNext() );
assertFalse( nodeAutoIndex.get( "nodeProp1", "nodeProp1Value2" ).hasNext() );
// node2 is updated
nodeAutoIndex.get( "nodeProp2", "nodeProp2Value2" ).getSingle() );
/*
* node3 is still there, despite its nodeProp1 property not being monitored
* any more because it was not touched, in contrast with node1.
*/
nodeAutoIndex.get( "nodeProp1", "nodeProp3Value" ).getSingle() );
// Finally, node4 is removed because the property was removed.
}
Caution
If you start the database with auto indexing enabled but different auto indexed properties
than the last run, then already auto-indexed properties will be deleted from the index when
a value is written to them (assuming the property isnt present in the new configuration).
Make sure that the monitored set is what you want before enabling the functionality.
572
Chapter35.Batch Insertion
Neo4j has a batch insertion facility intended for initial imports, which bypasses transactions and other
checks in favor of performance. This is useful when you have a big dataset that needs to be loaded
once.
Batch insertion is included in the neo4j-kernel <http://search.maven.org/#search|ga|1|neo4j-kernel>
component, which is part of all Neo4j distributions and editions.
Be aware of the following points when using batch insertion:
The intended use is for initial import of data.

Batch insertion is not thread safe.
Batch insertion is non-transactional.
Batch insertion will re-populate all existing indexes and indexes created during batch insertion on
shutdown.
Unless shutdown is successfully invoked at the end of the import, the database files will be corrupt.
Warning
Always perform batch insertion in a single thread (or use synchronization to make only
one thread at a time access the batch inserter) and invoke shutdown when finished.
573
Batch Insertion
35.1.Batch Inserter Examples

Creating a batch inserter is similar to how you normally create data in the database, but in this case
the low-level BatchInserter <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/unsafe/
batchinsert/BatchInserter.html> interface is used. As we have already pointed out, you cant have
multiple threads using the batch inserter concurrently without external synchronization.
Tip
The source code of the examples is found here: BatchInsertDocTest.java <https://

github.com/neo4j/neo4j/blob/2.0.0/community/kernel/src/test/java/examples/
BatchInsertDocTest.java>
To get hold of a BatchInseter, use BatchInserters <http://components.neo4j.org/neo4j/2.0.0/apidocs/
org/neo4j/unsafe/batchinsert/BatchInserters.html> and then go from there:
BatchInserter inserter = BatchInserters.inserter( "target/batchinserter-example", fileSystem );
Label personLabel = DynamicLabel.label( "Person" );
inserter.createDeferredSchemaIndex( personLabel ).on( "name" ).create();
Map<String, Object> properties = new HashMap<>();
properties.put( "name", "Mattias" );
long mattiasNode = inserter.createNode( properties, personLabel );
properties.put( "name", "Chris" );
long chrisNode = inserter.createNode( properties, personLabel );
RelationshipType knows = DynamicRelationshipType.withName( "KNOWS" );
// To set properties on the relationship, use a properties map
// instead of null as the last parameter.
inserter.createRelationship( mattiasNode, chrisNode, knows, null );
inserter.shutdown();
To gain good performance you probably want to set some configuration settings for the batch inserter.
Read Section22.11.3, Batch insert example for information on configuring a batch inserter. This is
how to start a batch inserter with configuration options:
Map<String, String> config = new HashMap<>();
BatchInserter inserter = BatchInserters.inserter(
"target/batchinserter-example-config", fileSystem, config );
// Insert data here ... and then shut down:
In case you have stored the configuration in a file, you can load it like this:
try ( InputStream input = fileSystem.openAsInputStream( new File( "target/batchinsert-config" ) ) )
{
Map<String, String> config = MapUtil.load( input );
BatchInserter inserter = BatchInserters.inserter(
"target/batchinserter-example-config", fileSystem, config );
// Insert data here ... and then shut down:
}
574
Batch Insertion
35.2.Batch Graph Database

In case you already have code for data import written against the normal Neo4j API, you could
consider using a batch inserter exposing that API.
Note
This will not perform as good as using the BatchInserter API directly.
Also be aware of the following:
Starting a transaction or invoking Transaction.finish()/close() or Transaction.success() will do
nothing.
Invoking the Transaction.failure() method will generate a NotInTransaction exception.
Node.delete() and Node.traverse() are not supported.
Relationship.delete() is not supported.
Event handlers and indexes are not supported.
GraphDatabaseService.getRelationshipTypes(), getAllNodes() and getAllRelationships() are not
supported.
With these precautions in mind, this is how to do it:
GraphDatabaseService batchDb =
BatchInserters.batchDatabase( "target/batchdb-example", fileSystem );
Label personLabel = DynamicLabel.label( "Person" );
Node mattiasNode = batchDb.createNode( personLabel );
mattiasNode.setProperty( "name", "Mattias" );
Node chrisNode = batchDb.createNode();
chrisNode.setProperty( "name", "Chris" );
chrisNode.addLabel( personLabel );
RelationshipType knows = DynamicRelationshipType.withName( "KNOWS" );
mattiasNode.createRelationshipTo( chrisNode, knows );
batchDb.shutdown();
Tip
The source code of the example is found here: BatchInsertDocTest.java <https://

github.com/neo4j/neo4j/blob/2.0.0/community/kernel/src/test/java/examples/
BatchInsertDocTest.java>
575
Batch Insertion
35.3.Index Batch Insertion

For general notes on batch insertion, see Chapter35, Batch Insertion.
Indexing during batch insertion is done using BatchInserterIndex <http://components.neo4j.org/
neo4j/2.0.0/apidocs/org/neo4j/unsafe/batchinsert/BatchInserterIndex.html> which are provided via
BatchInserterIndexProvider <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/unsafe/
batchinsert/BatchInserterIndexProvider.html>. An example:
BatchInserter inserter = BatchInserters.inserter( "target/neo4jdb-batchinsert" );
BatchInserterIndexProvider indexProvider =
new LuceneBatchInserterIndexProvider( inserter );
BatchInserterIndex actors =
indexProvider.nodeIndex( "actors", MapUtil.stringMap( "type", "exact" ) );
actors.setCacheCapacity( "name", 100000 );
Map<String, Object> properties = MapUtil.map( "name", "Keanu Reeves" );
long node = inserter.createNode( properties );
actors.add( node, properties );
//make the changes visible for reading, use this sparsely, requires IO!
actors.flush();
// Make sure to shut down the index provider as well
indexProvider.shutdown();
The configuration parameters are the same as mentioned in Section34.10, Configuration and fulltext
indexes.
35.3.1.Best practices
Here are some pointers to get the most performance out of BatchInserterIndex:
Try to avoid flushing <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/unsafe/
batchinsert/BatchInserterIndex.html#flush%28%29> too often because each flush will result in all
additions (since last flush) to be visible to the querying methods, and publishing those changes can
be a performance penalty.
Have (as big as possible) phases where one phase is either only writes or only reads, and dont
forget to flush after a write phase so that those changes becomes visible to the querying methods.
Enable caching <http://components.neo4j.org/neo4j/2.0.0/apidocs/org/neo4j/unsafe/batchinsert/
BatchInserterIndex.html#setCacheCapacity%28java.lang.String,%20int%29> for keys you know
youre going to do lookups for later on to increase performance significantly (though insertion
performance may degrade slightly).
Note
Changes to the index are available for reading first after they are flushed to disk. Thus,
for optimal performance, read and lookup operations should be kept to a minimum during
batchinsertion since they involve IO and impact speed negatively.
576
AppendixA.Manpages
The Neo4j Unix manual pages are included on the following pages.
neo4j
neo4j-installer
neo4j-shell
neo4j-backup
neo4j-arbiter
577
Manpages
Name
neo4j Neo4j Server control
Synopsis
neo4j <command>
DESCRIPTION
Neo4j is a graph database, perfect for working with highly connected data. The neo4j command is
used to control the Neo4j Server.
The preferred way to install Neo4j on Linux systems is by using prebuilt installation packages, but
theres also the possibility to use the neo4j-installer command to install or remove it as a system
service. For information regarding Windows, see below.
COMMANDS
console
Start the server as an application, running as a foreground process. Stop the server using CTRL-C.
start
Start server as daemon, running as a background process.
stop
Stops a running daemonized server.
restart
Restarts the server.
status
Current running state of the server.
info
Displays configuration information, such as the current NEO4J_HOME and CLASSPATH.
Usage - Windows
Neo4j.bat
Double-clicking on the Neo4j.bat script will start the server in a console. To quit, just press control-C
in the console window.
For installing the Neo4j Server as a service, use the Neo4jInstaller.bat command.
Neo4j.bat start - will start the Neo4j service
will start the Neo4j service if installed or a console
session otherwise.
Neo4j.bat stop - stop the Neo4j service if running
Neo4j.bat restart - restart the Neo4j service if installed
Neo4j.bat status - report on the status of the Neo4j service
returns RUNNING, STOPPED or NOT INSTALLED
FILES
Server configuration.
578
Manpages
conf/neo4j-wrapper.conf
Configuration for service wrapper.
conf/neo4j.properties
Tuning configuration for the database.
579
Manpages
Name
neo4j-installer Neo4j Server installation and removal
Synopsis
neo4j-installer <command>
DESCRIPTION
Neo4j is a graph database, perfect for working with highly connected data.
The preferred way to install Neo4j on Linux systems is by using prebuilt installation packages, but
theres also the possibility to use the neo4j-installer command to install or remove it as a system
service. For information regarding Windows, see below.
Use the neo4j command to control the Neo4j Server.
COMMANDS
install
Installs the server as a platform-appropriate system service.
remove
Uninstalls the system service.
Usage - Windows
To just control the Neo4j Server, use the Neo4j.bat command.
Neo4jInstaller.bat install/remove
Neo4j can be installed and run as a Windows Service, running without a console window. Youll need
to run the scripts with Administrator privileges. Just use the Neo4jInstaller.bat script with the proper
argument:
Neo4jInstaller.bat install - install as a Windows service
will install the service
Neo4jInstaller.bat remove - remove the Neo4j service
will stop and remove the Neo4j service
FILES
Server configuration.
conf/neo4j-wrapper.conf
conf/neo4j.properties
Tuning configuration for the database.
580
Manpages
Name
neo4j-shell a command-line tool for exploring and manipulating a graph database
Synopsis
neo4j-shell [REMOTE OPTIONS]
neo4j-shell [LOCAL OPTIONS]
DESCRIPTION
Neo4j shell is a command-line shell for running Cypher queries. Theres also commands to get
information about the database. In addition, you can browse the graph, much like how the Unix shell
along with commands like cd, ls and pwd can be used to browse your local file system. The shell can
connect directly to a graph database on the file system. To access local a local database used by other
processes, use the readonly mode.
REMOTE OPTIONS
-port PORT
Port of host to connect to (default: 1337).
-host HOST
Domain name or IP of host to connect to (default: localhost).
-name NAME
RMI name, i.e. rmi://<host>:<port>/<name> (default: shell).
-readonly
Access the database in read-only mode.
LOCAL OPTIONS
-path PATH
The path to the database directory. If there is no database at the location, a new one will e created.
-pid PID
Process ID to connect to.
-readonly
Access the database in read-only mode.
-c COMMAND
Command line to execute. After executing it the shell exits.
-file FILE
File to read and execute. After executing it the shell exits. If - is supplied as filename data is read
from stdin instead.
-config CONFIG
The path to the Neo4j configuration file to be used.
EXAMPLES
Examples for remote:
neo4j-shell
neo4j-shell -port 1337
neo4j-shell -host 192.168.1.234 -port 1337 -name shell
neo4j-shell -host localhost -readonly
581
Manpages
Examples for local:
neo4j-shell -path /path/to/db
neo4j-shell -path /path/to/db -config /path/to/neo4j.properties
neo4j-shell -path /path/to/db -readonly
582
Manpages
Name
neo4j-backup Neo4j Backup Tool
Synopsis
neo4j-backup -from SourceURI -to Directory
DESCRIPTION
A tool to perform live backups over the network from a running Neo4j graph database onto a local
filesystem. Backups can be either full or incremental. The first backup must be a full backup, after that
incremental backups can be performed.
The source(s) are given as URIs in a special format, the target is a filesystem location.
BACKUP TYPE
-full
copies the entire database to a directory.
-incremental
copies the changes that have taken place since the last full or incremental backup to an existing
backup store.
The backup tool will automatically detect whether it needs to do a full or an incremental backup.
SOURCE URI
Backup sources are given in the following format:
<running mode>://<host>[:<port>][,<host>[:<port>]]
Note that multiple hosts can be defined.
running mode
'single' or 'ha'. 'ha' is for instances in High Availability mode, 'single' is for standalone databases.
host
In single mode, the host of a source database; in ha mode, the cluster address of a cluster member.
Note that multiple hosts can be given when using High Availability mode.
port
In single mode, the port of a source database backup service; in ha mode, the port of a cluster
instance. If not given, the default value 6362 will be used for single mode, 5001 for HA
IMPORTANT
Backups can only be performed on databases which have the configuration parameter
enable_online_backup=true set. That will make the backup service available on the default port (6362).
To enable the backup service on a different port use for example enable_online_backup=port=9999
instead.
Usage - Windows
The Neo4jBackup.bat script is used in the same way.
EXAMPLES
Performing a backup the first time: create a blank directory and run the backup tool
583
Manpages
mkdir /mnt/backup/neo4j-backup
neo4j-backup -from single://192.168.1.34 -to /mnt/backup/neo4j-backup
Subsequent backups using the same _target_-directory will be incremental and therefore quick
neo4j-backup -from single://freja -to /mnt/backup/neo4j-backup
Performing a backup where the service is registered on a custom port
neo4j-backup -from single://freja:9999 -to /mnt/backup/neo4j-backup
Performing a backup from HA cluster, specifying only one cluster member
./neo4j-backup -from ha://oden:5002 -to /mnt/backup/neo4j-backup
Performing a backup from HA cluster, specifying two cluster members
./neo4j-backup -from ha://oden:5001,loke:5002 -to /mnt/backup/neo4j-backup
RESTORE FROM BACKUP

The Neo4j backups are fully functional databases. To use a backup, replace your database directory
with the backup.
584
Manpages
Name
neo4j-arbiter Neo4j Arbiter for High-Availability clusters
Synopsis
neo4j-arbiter <command>
DESCRIPTION
Neo4j Arbiter is a service that can help break ties in Neo4j clusters that have an even number of
cluster members.
COMMANDS
console
Start the server as an application, running as a foreground process. Stop the server using CTRL-C.
start
Start server as daemon, running as a background process.
stop
Stops a running daemonized server.
restart
Restarts a running server.
status
Current running state of the server
install
Installs the server as a platform-appropriate system service.
remove
Uninstalls the system service
FILES
conf/arbiter.cfg
Arbiter server configuration.
conf/arbiter-wrapper.cfg
585

Neo4j-Manual-2 0 0

Uploaded by

Copyright:

Available Formats

Neo4j-Manual-2 0 0

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Neo4j-Manual-2 0 0

Uploaded by

Copyright:

Available Formats

The Neo4j Manual v2.0.

The Neo4j Team neo4j.org <http://neo4j.org/>

The Neo4j Manual v2.0.0

by The Neo4j Team neo4j.org <http://neo4j.org/> www.neotechnology.com <http://www.neotechnology.com/>

What is a graph database?

License: Creative Commons 3.0

The Neo4j Manual v2.0.0

PartI, Introduction introducing graph database concepts and Neo4j.

true ACID transactions,

Chapter2.Graph Database Concepts

Graph Database Concepts

2.1.What is a Graph Database?

2.1.1.A Graph contains Nodes and Relationships

2.1.2.Relationships organize the Graph

2.1.3.Labels group the Nodes

Graph Database Concepts

2.1.4.Query a Graph with a Traversal

2.1.5.Indexes look-up Nodes or Relationships

Graph Database Concepts

2.1.6.Neo4j is a Graph Database

Graph Database Concepts

Graph Dat abase

Graph Database Concepts

2.2.Comparing Database Models

2.2.1.A Graph Database transforms a RDBMS

Figure2.2.Graph Database as RDBMS

2.2.2.A Graph Database elaborates a Key-Value Store

Graph Database Concepts

2.2.3.A Graph Database relates Column-Family

2.2.4.A Graph Database navigates a Document Store

Graph Database Concepts

D=Document, S=Subdocument, V=Value, D2/S2

= reference to subdocument in (other) document.

Figure2.6.Graph Database as Document Store

Chapter3.The Neo4j Graph Database

The Neo4j Graph Database

The Neo4j Graph Database

Relat ionship t ype

uniquely ident ified by

out going relat ionship

The Neo4j Graph Database

Using relationship direction and type

get who a person follows

outgoing follows relationships, depth one

get the followers of a person

incoming follows relationships, depth one

get who a person blocks

outgoing blocks relationships, depth one

get who a person is blocked by

incoming blocks relationships, depth one

The Neo4j Graph Database

can be an array of can be a

Property value types

32-bit IEEE 754 floating-point number

64-bit IEEE 754 floating-point number

The Neo4j Graph Database

16-bit unsigned integers representing

sequence of Unicode characters

The Neo4j Graph Database